Trésor-Info - Publications de la direction générale du Trésor - post-agnes-benassy-quere

Counter-cyclical unemployment insurance: why?

2023-02-21T00:00:00+01:00

According to INSEE, in the 4th quarter of 2022, 62% of companies in services (excluding transport), 65% in manufacturing and 81% in construction were still reporting hiring constraints. These are the highest levels ever recorded. Conversely, supply chains disruptions were on the decline. This diagnosis has been confirmed by the Banque de France survey in January 2023.

The decline in the unemployment rate over the years 2021 and 2022 (about -1 percentage point between 2021Q1 and 2022Q3, or about 200,000 fewer ILO unemployed over the period) shows that companies have been able to fill many vacancies with previously unemployed people. Over a longer period, the almost continuous decline in the unemployment rate since 2015 has been accompanied by a relatively modest increase in the proportion of job vacancies (Figure 1). However, with 7.3% unemployment (2022Q3 figure), we have reached a point where the increase in the proportion of job vacancies is no longer accompanied by a significant fall in unemployment: the Beveridge curve, which links the unemployment rate to the job vacancy rate, has become much steeper in France.

Figure 1

However, there is no real reason to leave it there. If we zoom out again by juxtaposing the Beveridge curves of different advanced countries, we can see that the unemployment rate below which the job vacancy rate increases is rather around 3-4% in Germany, the United Kingdom or the United States (Figure 2).

Figure 2

Sources : INSEE, Dares, Eurostat. DG Treasury calculations.

The premature steepening of the Beveridge curve in France betrays a relative inefficient matching process on the labour market: some unemployed people are unable to find a job while, at the same time, some employers cannot find the labour they need. Part of this mismatch between supply and demand is structural: the skills and geographical constraints of job seekers are not necessarily in line with the demand of the firms, at least in the short term. However, existing research also points to the role of the unemployment insurance system. In France, it provides benefits for up to 2 years before the age of 53 and 3 years from the age of 55, a relatively long period in international comparison (Figure 3).

Figure 3: Maximum duration of unemployment benefit in some European countries

Source: Panorama des systèmes d'assurance chômage en Europe, Unédic, January 2022; L'assurance chômage en Europe - Etude de 15 pays, Unédic, July 2019; "L'indemnisation du chômage en Suisse", Unédic, 2014; "L'indemnisation du chômage en Allemagne", Unédic, 2019; Agence pour le développement de l'emploi (ADEM) du Luxembourg; "L'indemnisation du chômage en Finlande", Unédic, June 2022; European Commission – Unemployment benefits in Portugal.

Incentives of the insured

Whatever its purpose (housing, automobile, occupational risks, illness, unemployment), the purpose of insurance is to protect an individual or an organisation against a risk. This protection comes up against the well-known problem of moral hazard: being insured modifies the behaviour of the individual or organisation. Insurers have developed strategies to protect themselves against moral hazard. For example, they impose a malus on drivers after an accident; they require security equipment from their policyholders (armoured doors); or they carry out random checks, for example on people who are off sick.

In the case of unemployment, the temporal limit of the benefit is the main mechanism to limit moral hazard, by encouraging the return to work, especially when the date of termination of the benefit is approaching. This mechanism has been empirically corroborated by Marinescu and Skandalis (2021) on the French case. Using a sample of 500,000 jobseekers who became unemployed between 2013 and 2017, the two researchers show that job search effort (measured by applications sent via an online platform) increases by 50% in the year preceding the end of the benefit. Therefore, a shortening of the benefit period would be likely to reduce the unemployment period. Also using French data, Le Barbanchon, Rathelot and Roulet (2019) estimate that a 10% shortening of the potential benefit period reduces the actual unemployment benefit period by an average of 3%, a figure comparable to that obtained in other European countries (see Roulet, 2018). According to the same studies, the time to return to work is reduced by 1% following a 10% reduction in the potential duration of benefits.

However, one can wonder whether a shorter benefit period could deteriorate the quality of the job found after the unemployment period, as the insured person would be encouraged to accept a job that is not in line with his or her qualifications and aspirations. Employers could also take advantage of this rush by offering lower wages. Although the economic literature is not unanimous on this point (Les droits rechargeables - UNÉDIC - October 2019), several studies carried out for France and other European countries highlight the negligible effects of extending the compensation period on the quality of employment: the wage recovered would not be higher, nor would the duration of employment (Le Barbanchon, 2016), the job would not be more stable (Fackler, Stegmaier and Weigt, 2019), the number of hours sought, the type of contract sought, or the maximum journey accepted between home and the workplace would not be affected either (Le Barbanchon, Rathelot and Roulet, 2019). In reality, two opposing effects seem to offset each other: on the one hand, an increase in the duration of compensation allows a form of selectivity on the part of jobseekers, and therefore access to a better job; on the other hand, a longer period spent unemployed reduces the prospects of hiring, which translates into a penalty on the quality of the job found.

Countercyclicality

When thinking about an optimal unemployment insurance system, one must also take into account the situation of the labour market, as shown by Landais, Michaillat and Saez (2018). If few firms are looking to recruit, a limited search effort by an unemployed person makes another more motivated unemployed person find a job more easily. Thus, when there are few vacancies in the market, the disincentive effects of insurance have limited impact on aggregate unemployment. Conversely, if, as is the case today, many companies are looking to recruit, then on the contrary, the job search of some does not affect the search of others, and unemployment decreases more rapidly when the search effort increases. Taking into account the situation of the labour market and its possible congestion, it is advisable to have a counter-cyclical unemployment insurance: more generous in the trough of the employment cycle, less so at the peak. Moreover, a countercyclical UI system better fulfils its insurance role, since it better protects the worker in periods when the probability of finding a job is low (Andersen, 2014).

In the United States and Canada, unemployment benefits increase in times of crisis and are paid for longer than in so-called "normal" times. In France, on the contrary, the parameters of unemployment insurance do not vary during the activity cycle. According to Cahuc, Carcillo and Landais, 2020, the compensation rules even tend to be slightly pro-cyclical. However, Schmieder, von Wachter and Bender (2012) show, on German data, that the disincentive effects of unemployment insurance to return to work are more limited during a recession, suggesting welfare gains associated with the countercyclical modulation of the potential duration of compensation.

Practical work

Should the potential duration of compensation, the amount of compensation (replacement rate) or the eligibility conditions be modulated? The choice has implications in terms of redistribution and incentives. In particular, modulating the replacement rate over the cycle could involve affecting financially constrained workers in troughs (Chetty, 2008). Conversely, modulating the potential duration of benefits via the conversion rate makes it possible to affect all workers, including those who have accumulated savings during their period of employment, and without any noticeable effect on the quality of the job found (see above). A modulation of the maximum duration of compensation alone would have a slower effect and would risk penalising certain categories of unemployed more heavily (those with the longest entitlements). Finally, a modulation of the eligibility conditions would have effects that are difficult to control given the length of the period of employment and could deprive certain workers of their rights, in particular young people.

The second issue is the choice of the trigger variable. This variable must be able to reflect the labour market situation faced by a jobseeker in a quick, objective, understandable and robust way with respect to possible labour market developments. On the basis of these different criteria, the unemployment rate must be prefered to indicators of inflows and outflows from unemployment, the job vacancy rate and the output gap (between GDP and potential GDP). The unemployment rate is calculated by an independent institution - INSEE - based on an international definition given by the International Labour Office. It is little subject to revision, well known to the general public and its link to the labour market is clear to everyone.

Once the trigger variable has been chosen, it is necessary to find the appropriate thresholds. Decree No. 2023-33 of 26 January 2023, applicable from 1 February to 31 December 2023, provides for a 25% reduction in the potential duration of compensation when, for three consecutive quarters, the unemployment rate remains below 9% and does not vary by more than 0.8 percentage point quarter on quarter. An end-of-right supplement is granted when unemployment exceeds 9% during a quarter, or if the unemployment rate changes by more than 0.8 percentage point qoq. This second criterion is introduced in order to be able to react quickly to a sudden deterioration in the economic situation, even if the unemployment rate remains low, without triggering too frequent changes in the parameters of the UI system. The decree provides for a minimum duration of compensation of 6 months for eligible persons who have paid contributions for at least 6 months, which maintains the insurance function of the scheme for all eligible workers, even during the peak of the cycle, and allows for a good match on the labour market. The conditions of eligibility and the amount of the monthly allowance remain unchanged.

Figure 4 shows what the potential duration of benefits would have been over the period 2003-2022 if this modulation had been applied. The graph is in a green zone when the labour market situation is favourable and in a red zone when it is deteriorated. If the compensation rules contained in the decree had been in place since 2003, the unemployed would have had a top-up from the global crisis of 2009 until 2018 and again, briefly, in 2020. Thus, over twenty years, the combination of the two criteria (level and variation of the unemployment rate) would have made it possible to identify the two main crises, the 2009 crisis and the Covid crisis, and only these two episodes.

Figure 4

Reading: In 2018Q1, under the modulation rules introduced by the new decree, the cycle would have been considered unfavourable: the unemployment rate was above 9%. In 2019Q2, it would have been considered favourable because of the prolonged fall in unemployment to below 9% (with no quarterly change of more than 0.8 percentage points over the period). From 2020Q3 onwards, it would have been considered unfavourable again for three quarters, due to the sudden increase in the unemployment rate in 2020Q2.

Source: DG Treasury, based on INSEE data.

Hence, the decree fulfils the specifications set out in the academic literature, bringing the French scheme closer to the "optimal" trade-off between insurance and incentives, which varies over the cycle. Modulation according to a clear rule gives visibility to workers and will allow balancing the benefit system over the cycle, contributing both to macroeconomic stabilisation and to the sustainability of public finances. Maximising employment at the top of the cycle also responds to the needs of the firms. In addition to UI contributions, the wealth created will provide tax resources for the general government.

Read more:

>> Version française : Assurance-chômage contra-cyclique : pourquoi ?

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

A price-wage loop on the Christmas tree?

2022-12-29T00:00:00+01:00

As the year draws to a close, inflation is still high and could remain so for the next few months according to the latest forecasts from INSEE and the Banque de France. While over the last 30 years (1991-2021), consumer prices have increased by an average of 1.5% per year in France, the year-on-year increase reached 6.2% in October and November 2022 (figure 1).

Figure 1. Year-on-year change in the consumer price index in France, 1991-2022, in %

Source: INSEE.

The surge in energy and commodity prices started before Russia invaded Ukraine. During 2022, their contribution in consumer price inflation has been the greatest. The rise in input costs was also passed on to manufacturing prices, particularly in food industries, and gradually in services (figure 2). As companies are each other's customers, the rise in prices for some leads to a rise in prices for others and vice versa. This is the price-price loop.

Figure 2. Contributions to year-on-year consumer price inflation in France, 2021-2022

(in percentage points)

Interpretation: in March 2022, energy contributed 2.3 points to a CPI inflation of 4.5% over one year; in November, energy contributed 1.6 points to a total of 6.2%.

Source: INSEE, calculations by DG Treasury.

In recent months, the price of imported energy and raw materials has fallen back to a level close to that observed before the war in Ukraine (see INSEE, 16 December 2022), which should ease downstream prices. In 2023, one of the main drivers of inflation should in fact be wage growth, which could exceed 6% over one year according to the Banque de France (figure 3). This rise in wages will in turn feed the rise in prices. This is the price-wage loop.

Figure 3: Average nominal earnings per head in France, market sector, year-on-year % change

Source: Insee, and Banque de France forecasts (December 2022). The average wage is corrected for the effect of short-time work.

Some observers, considering that there is no evidence of a wage-price loop, advocate a generalised indexation of wages. They refer to a study by the International Monetary Fund according to which a period of wage catch-up does not necessarily lead to a runaway price-wage spiral. However, the IMF sample specifically excludes periods of wage indexation, with the exception of the US in the 1970s, an episode that the IMF points to as an example of a runaway. The study shows in fact that after an inflationary surge, nominal wages gradually "catch up" with prices, without the need for explicit indexation. More fundamentally, the complete absence of a wage-price loop would mean that companies would be able to absorb all the increase in their costs without passing it on in their prices, or that employees would accept to see the real value of their wages erode with inflation without any wage compensation. In reality, we are now seeing inflation spread to non-energy prices and wages. The real question is therefore not whether or not there is a price-wage loop, but how strong it is, which will determine the strength of inflation in the coming months.

Wage de-indexation

In the 1970s, general wage indexation was a powerful force in the price-wage loop: when consumer prices rose by 1%, wages automatically rose by 1%, which increased companies' costs in proportion to their wage bill, pushing up prices, etc. At the time, productivity gains had allowed wage increases to exceed inflation: thanks to productivity gains, real wages had been able to grow without the real margins of companies decreasing, as the Governor of the Banque de France, François Villeroy de Galhau, recently reminded us during a speech at the Toulouse School of Economics. Unemployment rose sharply during this period, from 3.6% in 1975 to 7.1% in 1982.

During this period, the chase between wages and prices had fuelled inflation. Between 1973 and 1982, annual inflation averaged 11.2%. It was only from 1983 onwards that inflation started to fall in France, under the combined effect of a restrictive monetary policy, a commitment not to devalue any more, a restrictive fiscal policy (la rigueur), and a ban on indexation clauses (see Mojon and Pereira da Silva, 2019).

Since then, wages are no longer automatically indexed in France, with the exception of the minimum wage (Smic), which is revalued each year on 1 January according to (1) the increase in the consumer price index excluding tobacco of the first quintile of households in terms of standard of living, and (2) half the increase in the purchasing power of the basic hourly wage of manual and clerical workers (see the Report of the Smic Expert Group, 22 November 2022). When inflation is high, as was the case this year, the minimum wage is also revalued during the year.

Increases in the minimum wage automatically lead to the revaluation of certain branch minimum wages which would otherwise be "caught up" by the minimum wage. More generally, hikes in the minimum wage increase both the probability of revaluation and the extent of revaluations of all wages during negotiations within companies, in addition to the effect of inflation itself (see in particular the recent estimate on micro data by Gautier, Roux and Suarez Castillo, 2022).

At the macro level, we do observe an increase in wages when consumer prices rise. In the Opale model, which is used by the DG Treasury to make macroeconomic forecasts, a 10% rise in the consumer price index leads, all other things being equal, to a cumulative increase in the average nominal wage per capita of 6.4% in the first year, 8% in the second year and 10% in the long term. From this point of view, the increase in the average wage per capita forecast by the Banque de France for 2022 and 2023 (respectively +3.8% and +6.2% after correction for the effect of partial activity) is consistent with inflation.

According to DARES, the basic monthly wage, which does not include bonuses, would have increased by an average of 3.7% between the third quarter of 2021 and the third quarter of 2022, with a more marked increase at the bottom of the wage scale, and therefore a narrowing of the pay scale (figure 4). We see here a major advantage of the absence of automatic wage indexation, with the exception of the minimum wage: the terms of trade shock affects workers differently depending on their position in the wage scale, with those at the lower end being protected. In addition, the absence of automatic indexation allows companies to modulate revaluations according to their exposure to the energy shock and their capacity to pass on the increase in costs to their customers. They can also spread revaluations over time, depending on the evolution of their margins. Conversely, a uniform revaluation would put some companies in difficulty in the short term.

Figure 4: Evolution of the basic monthly wage in France between 2021-Q3 and 2022-Q3, in %.

Source: Dares. Year-on-year variations in current euros.

Tariff shields

Faced with the energy shock, the strategy of the French government was to spread the effects over time via "tariff shields": a retail price control for electricity and gas to which was added the fuel discount until December 2022. The idea was to give households and businesses time to organise themselves in the face of these price increases in order to reduce their vulnerability; but also to limit the volatility of relative prices in order to reduce the negative effects on the economy. The evolution of energy prices during the year 2022 validates this approach for the moment. For example, the market price of natural gas for Europeans increased by a factor of 10 between January 2021 and August 2022, but by "only" 3 between January 2021 and October 2022 (source: Fred).

However, this policy means that the fall in energy prices is not immediately reflected in the consumer price index either, unlike in other countries, for example Spain, where inflation has fallen from its peak of 10.8% in July. The impact of tariff shields on prices in the long term will depend on how the smoothed inflation profile affects expectations and thus wage formation. It will also depend on energy efficiency efforts that may ease prices, especially for electricity, whose market is not globalised.

In a closed macroeconomic framework, Cepremap has estimated that the tariff shields are likely to have reduced inflation by about 1.1 percentage points and supported activity by about 1.75 percentage points in 2022. According to this study, general wage indexation would have led to higher inflation in 2022, with lower GDP and higher fiscal cost. The "heterogeneous agents" framework also allows for an analysis of the impact of different policies on inequality. The authors find that a generalised indexation would have done less well from this point of view than the shields, also taking into account the indexation of social transfers (which constitute nearly 70% of the income for the first decile of the standard of living), and even before accounting for the indexation of the minimum wage.

Purchasing power and the labor share

The terms of trade shock suffered by the French economy has been evaluated by the DG Treasury at 3% of GDP in 2022 compared to 2019, fosusing on the surge in imported energy prices themselves, before any pass-through or adjustment mechanism. The downward revision of growth for 2022 and 2023 since the outbreak of the war is consistent with what macroeconomic models such as Mésange predict after an exogenous deterioration in the terms of trade. The shock thus causes a reduction in value added, to be distributed among economic agents.

France differs from the other advanced countries by a stable labor share over the long term. In order to maintain a stable distribution of value added during the energy crisis, the fall in value added (compared to a non-crisis scenario) must be distributed proportionally between the two production factors: the real remuneration of labour and capital must not grow faster than GDP, the growth of which has been revised downwards.

In its latest Economic Outlook, INSEE warns of a fall in margin rates in industry excluding energy, construction, transports and other market services in 2022, due to higher input prices. For its part, employment dynamics have so far compensated for the decline in average real wages. If we add to this the indexation of social benefits and the exceptional vouchers, purchasing power has been largely preserved. By the end of 2022, according to the OECD, household purchasing power would be 2% above the level at the end of 2019, while it would be 4% below its pre-Covid level in Germany and the UK (figure 5).

Figure 5: Changes in household purchasing power since Q4 2019

Source: OECD Economic Outlook (November 2022). Purchasing power is calculated as nominal disposable income adjusted for the private consumption deflator; net disposable income for Germany, gross for France and the UK.

The evolution of employment is now a major difference compared to the 1970s: in the third quarter of 2022, employment in France was 3.6% above its level in the fourth quarter of 2019, while GDP was only 1.1% above its pre-crisis level. Therefore, productivity gains have weakened, but employment has been buoyant whereas it had declined in the 1970s. While the government will have to gradually withdraw its support, employment and investment will be the key variables to alleviate the impact of this energy crisis on our economies.

***

>> Version française : Une boucle prix-salaires sur le sapin 2022 ?

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Looking for potential growth

2022-10-17T00:00:00+02:00

Following Richard and Peggy Musgrave, economists have become accustomed to grouping economic policies into three categories: those that raise the population's standard of living in the long term (structural policies), those that seek to limit its instability in the short to medium term (cyclical policies) and, finally, those that attempt to distribute the wealth produced more harmoniously than the market would spontaneously do (redistributive policies).

To differentiate between the first two categories, we use the concept of potential GDP, which can be defined as "the volume of production of goods and services that an economy can sustainably achieve by fully utilising its capacity, but without creating inflationary pressures" (Herlin and Gatier, 2017).

Like other macroeconomic concepts such as the structural unemployment rate, the neutral interest rate or the equilibrium exchange rate, potential growth has the major flaw of not being observable. It is estimated in two main ways (see Lequien and Montaut, 2014):

Either, from a modelling of output as a function of capital, labour and productivity. In this case, assumptions must be made about capital accumulation, changes in the labour force and working hours, changes in the unemployment rate and the rate of productivity growth;
Or, using a statistical method, by smoothing the fluctuations of GDP over a sufficiently long period, making the hypothesis that the observed GDP cannot deviate durably from the potential GDP.

The two methods can be combined. Thus, both French Treasury and the European Commission rely on a production function but insert potential employment and potential productivity, which are obtained by the smoothing method, while the potential capital stock follows the actual observed investment, corrected for capital depreciation (see Herlin and Gatier, 2017).

It should be noted straight away that, contrary to what the vocabulary might suggest, potential GDP is not an unsurpassable ceiling: GDP can exceed potential GDP when the economy is at the top of the cycle, as in 1990, 2000 or 2007 (Figure 1).

Figure 1. GDP and potential GDP of France according to the OECD, in constant 2014 € billion

Source: OECD, Economic Outlook, June 2022.

The gap between actual and potential GDP, known as the output gap, plays a major role in setting monetary and fiscal policy:

Monetary policy: a positive output gap in principle signals higher inflation, so that the central bank should tighten its policy; the output gap thus appears explicitly in the Taylor rule, which relates the short-term interest rate to current inflation and the output gap.
Fiscal policy: a positive output gap means that firms cannot produce more without increasing their unit costs. The economy is therefore not in a situation of Keynesian unemployment. In this configuration, a tightening of fiscal policy has little impact on output and can help limit inflationary pressures.

Potential GDP plays an explicit role in European fiscal rules through the division of the budget balance into two components: the cyclical balance and the so-called "structural" balance, i.e. corrected for the cycle and for one-off expenditures and revenues. When actual GDP is lower than potential GDP, as is the case today in France according to the OECD (Figure 1), part of the budget deficit is due to this negative output gap: it will automatically be absorbed when GDP returns to its potential level because tax revenues will recover. On the other hand, the structural part of the deficit will not be absorbed on its own: a discretionary decision to reduce expenditure or increase taxation will be necessary to bring the balance back to equilibrium. The preventive arm of the Stability and Growth Pact thus prescribes a reduction in the structural deficit of at least 0.5 points of GDP per year when the output gap is between -1.5 and +1.5%, until a structural balance "close to equilibrium" is reached (see the Vade-mecum of the Stability Pact, 2019 edition)

Hours worked

To understand potential GDP from the factors of production, the simplest way is to start with the following breakdown:

GDP = hours worked x GDP per hour worked

The number of hours worked is the product of the population of working age (15-64 years according to the international convention), the participation rate (number of active persons in relation to the population of working age), the proportion of active persons who are in employment (1 - unemployment rate) and the average duration of work (including part-time work). Analysing these different components allows us to anticipate the evolution of hours worked in the coming years.

For example, in France, over the period 2015-2019, hours worked increased by an average of 0.7% per year (Figure 2). This reflects an increase in the participation rate, the proportion of employed persons and the hours per employed person, while the population aged 15-64 started to decline. Hours worked increased at a much slower rate in the years 2010 to 2014 (less than 0.1% per year), partly because the proportion of employed persons was declining in the wake of the global financial crisis.

Figure 2. Average growth rate of hours worked in France between 2010 and 2019

Source: INSEE. Employment survey, Eurostat. Calculations by DG Treasury.

The latest INSEE projections anticipate a quasi-stability of the active population over the period 2023-2027. However, these projections do not take into account the positive effect on the participation rate of reforms that contribute to increasing the labour supply. By way of illustration, according to the calculations of the French Treasury, an increase of two years in the age of entitlement to retirement, at a rate of 3 months per year from 2023 onwards, would increase the active population by approximately 0.9% over a 5-year period, and 1.4% until completion of the reform.

The continuation of the decline in the unemployment rate observed since 2015 will depend on employment policies, but also on the evolution of productivity (see below) and, of course, on the macroeconomic situation. A continuation of the trend observed over the period 2015-19 would be consistent with developments elsewhere in Europe, a convergence towards lower levels of unemployment is now being discernible (Figure 3). All other things being equal, a fall in the structural unemployment rate of 2 percentage points over 5 years, or 0.4 pp per year, would roughly translate into a 0.4 pp increase in the growth rate of hours worked over this period.

Figure 3: Unemployment rates in different European countries, 2014-2022

As a % of the labour force

Source: OECD.

Productivity

The downward trend in the unemployment rate is mechanically accompanied by a slowdown in productivity per worker, as people who find a job have, temporarily or not, a slightly lower productivity than the average of those already in employment. According to the Dares Acemo-Covid survey, the main measure taken by companies to overcome recruitment difficulties would be to change the profile of the people recruited, particularly in terms of skills. The skill mismatch could become more pronounced as unemployment falls. Hiring difficulties, which are particularly high in 2022, may also lead companies to retain labour even in periods of under-activity.

The path of the unemployment rate over the period 2022-2027 is naturally very uncertain. However, it should be remembered here that GDP is the product of hours worked and productivity per hour (see above). To the extent that lower unemployment is at least temporarily accompanied by less dynamic productivity, an error in the evolution of unemployment would only partially affect potential growth (Figure 4).

Figure 4: The potential GDP factory

The potential growth assumption underlying the Economic, Social and Financial Report annexed to the 2023 Budget is 1.35% per year, of which about one-third is the result of growth in hours worked and the remaining two-thirds of productivity growth. The 2018-2022 Public Finance Programming Act already assumed potential growth of 1.35% in 2022, albeit with a smaller contribution from labour and a larger one from productivity (Table 1). In the meantime, employment has been more dynamic than expected, suggesting a growth that is richer in jobs (but conversely poorer in productivity) than in the past.

Table 1. Potential growth scenario of the French government

% per year	2022 (pre-Covid estimate) a	2023-2027 (October 2022 estimate) b	Révision b-a
Potential growth	1.35	1.35	0
Contributions:
Hours worked	0.2 à 0.3	0.4 à 0.5	+0.2
Labour productivity	1.1 à 1.2	0.9 à 1.0	-0.2
Of which: capital deepening	0.5	0.5	0
total factor productivity	0.6 à 0.7	0.4 à 0.5	-0.2

Sources : RESF 2022 and 2023.

The evolution of labour productivity is itself decomposed into two contributions:

Capital accumulation per worker
Growth in total factor productivity (TFP), also known as the combined efficiency of labour and capital.

While capital accumulation per worker has not weakened in recent decades, TFP has decelerated in advanced economies, France being no exception (Bergeaud, Cette and Lecat, 2020; Conseil national de productivité, 2019).

Long Covid for potential growth?

During the Covid crisis, most forecasters revised downwards their estimates of the level or growth rate of potential GDP, due to the temporary decline in investment (and thus, capital accumulation), but also to the experience of past crises that were almost always followed by a slowdown in TFP (see Herlin and Gatier, 2017). In the RESF 2021 multi-annual scenario (published in October 2020), a permanent loss in potential GDP of just over 2 percentage points was assumed. The very good performance of productive investment and employment, the resilience of the productive fabric, the absence of a proven zombification of the economy and, finally, the acceleration of the digitalisation of the economy (including the spread of teleworking) then led the Government, like other forecasters (Banque de France, OFCE in particular), to revise downwards the loss of potential GDP linked to the pandemic: it was estimated at -1 ¾ points in the RESF 2022, then at -¾ point in the RESF 2023 (Figure 5).

It must be said that unlike several past crises, notably the 2008 crisis, the Covid crisis was perfectly exogenous to the economy: as long as the production factors – capital and labour – had been preserved, a return to the growth path close to that of the pre-crisis period was now possible. As early as 2021, Fernald and Li (2021) predicted that the health crisis would not have a lasting impact on potential growth in the United States, and they went even further by judging that there would be no impact either.

Figure 5 : different estimates of potential GDP for France

Sources: INSEE, OECD, IMF, European Commission, RESF 2019 to 2022, DG Treasury calculations.

The estimate of potential growth varies according to the experts. Using a method similar to that of the government, the Independent Fiscal Council estimates it at 1.05% per year, due to a more conservative hypothesis on the effects of future reforms on the labour market and on productivity gains. For its part, the European Commission obtains a figure similar to that of the Government for the period 2022-2023, and lower thereafter (but without taking into account future labour market and pension reforms). The Government's estimate is comparable to that of the IMF in its April 2022 publication, or to that of the OFCE in its July forecast (see Table 2).

Table 2. Different estimates of potential growth post pandemic for France

(% per year, average 2022-2027)

	Date	% per year
Government	October 2022	1.35
OFCE	July 2022	1.3
Independent fiscal council	July 2022	1.05
European Commission*	May 2022	1.1
International Monetary Fund	April 2022	1.38

* 2022-2026. Source: French Treasury calculations.

What will be the impact of the energy crisis on potential growth? In the presence of a deterioration in the terms of trade, consumption is expected to be reduced but not necessarily production: higher import prices force people to produce more to maintain the same level of consumption, or to consume less for the same level of production (Kehoe and Ruhl, 2008).

However, the energy crisis is not just a terms of trade shock. Russia's invasion of Ukraine, but also concerns about Taiwan, could lead to a reorganisation of international value chains to limit supply risks. The counterpart would be higher costs, but also temporarily higher investments, with an uncertain net effect. Similarly, the ecological transition will have an uncertain effect on potential GDP, even if in the long term it will allow us to benefit from cheaper energy.

***

>> Version française : Chacun cherche sa croissance potentielle

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Unfair inflation

2022-09-19T00:00:00+02:00

Not everyone has the same consumption basket. You can calculate your own index using the INSEE simulator.

In its June 2022 Economic Outlook, INSEE estimated that, because of the lower weight of heating and food in their budget, young people fare better than older households. Similarly, residents of large cities with public transport are better off than rural households, while households in the lowest 10% are more strongly affected by inflation. These differences in consumption patterns tend to offset each other over time: on average over 2015-2021, inflation differences are negligible. If energy prices were to fall, older, rural and low-income households would be the first to benefit. In the short term, however, rising prices constrain the consumption of those who have no accumulated savings. For them, maintaining expenditure in value terms leads to a drop in the volume consumed.

In its latest publication on the subject, INSEE estimates that the rise in energy prices between the second quarter of 2021 and the second quarter of 2022 contributed 3.1 percentage points to the observed inflation (5.3%). The tenth of the population most affected is not the poorest one (3.3 percentage points) but the third one: households between the poorest 20% and the poorest 30% (3.5 percentage points). At the other end of the distribution, the contribution of energy to inflation would be only 2.7 points for the wealthiest 10% of households. In the absence of price controls (freezing of the regulated retail gas price, small increase in the regulated retail electricity price, fuel discount), the impact of the energy shock would have been much more disparate, peaking at 7% for the third tenth (Figure 1).

Figure 1. Effects of higher energy prices on inflation, by tenths of living standards

Q2 2021 to Q2 2022, in %

Interpretation: between the 2nd quarter of 2021 and the 2nd quarter of 2022, the rise in energy prices led to an increase in the cost of the consumption basket of 6.8% before price controls for the first tenth of households (the poorest 10%); an effect reduced to 3.3% due to price controls (gas, electricity and fuel).

Source: INSEE (2022).

The International Monetary Fund has calculated the impact of the rise in energy prices on the cost of living for the lowest 20% of households and for the highest 20%, in various European countries. The impact is very different from one country to another, both in terms of the average (for all households) and in terms of dispersion (inter-quintile range). France, along with Hungary and Finland, is the country where households are least affected on average and where, moreover, the incidence is closest between the first and last quintile. In contrast, the effect of higher energy prices is particularly marked in the UK and even more so in Estonia, two countries where the effect is also very different across income quintiles. Using a different method, the OFCE obtains qualitatively similar results.

Breaking the piggy banks

At the aggregate level, French households over-saved during the Covid crisis: according to the Banque de France, their financial savings exceeded their pre-crisis trend by more than €150bn in aggregate between the first quarter of 2020 and the first quarter of 2022, in cumulative terms. Assuming an average inflation of 5.5% in 2022 (Consensus Forecasts of August 2022 and INSEE forecast of June 2022, the French Stability Programme 2022 is assuming 5%), the additional cost of consumption this year would be around €70bn for households, i.e. less than half of the oversaving accumulated during the Covid crisis. Indeed, final household consumption (excluding individualisable public consumption) was €1,265bn in 2021.

While all deciles saved money during the Covid crisis, wealthy households did so more than others. According to the Conseil d'analyse économique (CAE), the median household would have fully consumed its 'Covid over-savings' by the end of 2021, its savings having then returned to their pre-crisis trend. But the lowest 10% would have drawn on their Covid savings to cope with the rise in prices: in June 2022, their savings stock would have been around 5% below the level that would have prevailed if the pre-pandemic trend had continued. However, the CAE does not detect any increase in financial insecurity as measured by the share of households with a bank account in deficit at the end of the month.

Changes in income

The final impact of price increases on the standard of living of households depends on the evolution of their income. In particular, we must take into account the indexation of certain incomes and targeted transfers to some categories of households. Table 1 below summarises the measures taken for the year 2022 in the case of France, following the Law of 16 August 2022 on emergency measures to protect purchasing power. Excluding permanent revaluations, but including price controls and the fuel discount, Bruegel calculates that France is among the countries that have provided the most support for household purchasing power since the start of the energy crisis.

Table 1. Household income support measures in France for the year 2022

Schemes	Measures
Social transfers
Inflation allowance	€100 paid at the end of 2021/beginning of 2022 to 39 million individuals with cumulated gross income from 1 January to 31 October 2021 not exceeding 26 000€.
Exceptional back-to-school bonus	€100 per household receiving minimum social benefits, housing benefits and/or in-work benefit, + €50 per dependent child in the household. Payment in September.
Value-sharing bonus (formerly exceptional purchasing power bonus)	Maximum annual amount of this tax-free bonus is doubled: up to €6,000 tax-free per employee if the company has a profit-sharing agreement. Exemption from income tax and social security contributions for employees earning less than 3 Smic (minimum wage) until the end of 2023.
Permanent revaluations
Basic retirement and disability pensions	+1.1% in January, +4% in July
Family allowances, minimum social benefits	+1.4% in April, +4% in July
In-work benefit	+4% in July
Personalized housing assistance (APL)	+3.5% in July
Revaluation of the minimum wage	+0.9% on 1/1, +2.65% on 1/5, +2.01% on 1/8, i.e. +8% de from Sep. 21 à to Sep. 22.
Revaluation of the civil service point	+3.5% on July 1^st
Lower contributions
Social security contributions for the self-employed	Cut in contributions (about €550 for net income close to the minimum wage)

Comparing Figure 1 and Table 1, it is clear that low-income households and older households are well targeted by income support measures: this targeting compensates for the greater impact of higher energy and food prices on these categories of households.

For low-income workers, the minimum wage will have risen by 8% between 1 September 2021 and 1 September 2022, for an annual increase in consumer prices of around 6%: even before taking into account the various bonuses and transfers mentioned in Table 1, their purchasing power will have increased. Wage increases are traditionally more gradual when moving away from the Smic, even if indexation of the Smic accelerates the rise in other wages (Gautier, Roux and Suarez Castillo, 2019). According to the Banque de France, sectoral wages negotiated in the last quarter of 2021 and the first quarter of 2022 were on the rise by 3% on average. The Dares estimates the increase in the basic hourly wage of manual and clerical workers between June 2021 and June 2022 at +3.5%. In the DG Treasury's forecasting model Opale, the rise in prices in year N leads to wage increases distributed between year N and year N+1.

Household incomes are also supported by job creation. Despite the energy crisis, 170,000 jobs have already been created in the private sector by 2022 Q2. Combined with wage increases, indexation and support measures, this employment dynamic means that purchasing power should remain stable in 2022.

The cost of war

Central banks are used to saying that inflation increases inequality because the incomes of modest households are less indexed to prices than those of wealthy households. This observation must be qualified in France, where minimum social benefits, pensions and the minimum wage are indexed to prices. In 2017, microsimulations carried out by INSEE concluded that a one-point increase in inflation had a small impact (-0.1%) on the standard of living of the first tenth of the standard of living (the 10% most modest) two years after the shock, the time profile depending on the indexation schedules. The study found a maximum impact (-0.6%) for the top 10% of households. However, the simulated price increase was a uniform increase, with all consumption items increasing in the same proportions. Furthermore, income from assets was assumed not to be indexed. However, some asset incomes - rents, regulated savings acounts in particular - are indexed, while others, such as listed shares, may see their value increase or decrease depending on the origin of inflation and the reaction of the central bank. On average, however, inflation is unfavourable to savers because nominal returns increase less than prices. On the contrary, it is favourable to indebted households insofar as, in France in particular, the majority of mortgages carry a fixed rate.

In France, household purchasing power increased by 5.8% cumulatively over the period 2017-2019. It remained stable in 2020 (+0.4%) and increased by 2.2% in 2021. The massive government support during the Covid crisis and now during the energy crisis is costly for public finances. If the increase in the cost of imported energy is permanent, then households and companies will have to gradually distribute the cost among themselves, with a negative impact on purchasing power. This is the real cost of the war for Europeans.

***

>> Version française : Injuste inflation

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Public finances: inflation that pays?

2022-07-12T00:00:00+02:00

It is often said that inflation is good for public finances, with essentially two arguments. The first is that a rise in consumer prices mechanically raises VAT revenues; a rise in wages mechanically raises the social contributions paid; and so on for each levy raised on a nominal base. The second argument is that government debt is fixed in euros (€2800bn in France at end-2021): dividing this debt by a rising nominal GDP, the ratio decreases. In other words, it becomes easier to repay a debt that has not changed.

These two arguments are not wrong, but they forget important elements, notably the origin of inflation. In the current context, the money we pay to oil and gas foreign suppliers will sooner or later run out of government coffers.

Primary balance

Let's start by looking at the impact of inflation on the primary balance, i.e. excluding interest payments. This balance is usually measured as a proportion of nominal GDP. At first sight, inflation increases revenues, expenditures and the denominator (GDP) in parallel, so that the ratio of the balance to GDP should not depend on prices. Why, then, do we have in mind that rising prices will reduce the deficit in the short run? The main reason is the only gradual indexation of expenditure.

When prices rise, tax revenues can increase faster than government expenditure. For example, if consumer prices rise, VAT revenues mechanically increase for a given volume of consumption. Similarly, rising wages increase the social security contributions collected, while rising corporate profits lead to additional corporate income tax revenues.

Some public expenditure is also indexed to prices: pensions, family allowances, housing allowances, minimum income, etc. are indexed within less than a year in France. Each additional percentage point of inflation automatically increases these social expenditures by almost €5 billion. However, the indexation periods for other expenditure are variable, for example for procurement contracts, or for civil service salaries. In the short term, inflation can therefore reduce the public deficit, but the effect will quickly dissipate.

However, so far we have considered all quantities as given: the tax bases (consumption, income) as well as expenditure are fixed in real terms. In fact, all of this also varies!

On the expenditure side, the government can decide on exceptional expenditure to protect the purchasing power of households and limit the impact of the increase in input costs for companies. In France, the inflation allowance, the exceptional reinforcement of the energy voucher, the tariff shields on gas and electricity and the aid granted to the most exposed companies are all discretionary expenditure caused by inflation and which increase the deficit (Table 1).

Table 1. Discretionary measures decided by the French government, Dec. 2021 to May 2022

Lump sum transfers to households

Tariff shields

Aid to businesses

Inflation allowance: €100 paid at the end of 2021 and beginning of 2022 to 38 million individuals.

Exceptional energy voucher: €100 paid to 5.8 million households in December 2021.

Exceptional back-to-school aid.

Early revaluation of pensions and benefits.

Revaluation of the civil service point

Electricity: increase in the regulated tariff limited to 4% in February 2022, extension of the regulated quota of nuclear power extended by EDF to retailers (ARENH).

Gas: freezing of the regulated tariff from October 2021.

Fuel: 15c€/l exl. VAT discount at the pump from April 2022.

Subsidies for energy-intensive companies.

Source: Budgetary documents.

On the revenue side, tax bases in real terms move in a direction that depends on the origin of inflation. If, as was the case at the beginning of the economic recovery after the Covid crisis, inflation is linked to a rapid recovery in demand, then the tax bases increase not only in value but also in real terms. But if, as now, inflation is driven by higher prices for imported energy and raw materials, then prices rise but GDP, employment and consumption are negatively affected (the economy becomes poorer). Table 2 compares the effect of a positive shock on world demand to that of an increase in oil prices, according to the Mesange macroeconomic model. In both cases, prices increase. But the primary balance improves in the former case while it deteriorates in the latter, due to a decrease in all tax bases in real terms.

Table 2. Impact of a demand shock and a supply shock on GDP, prices and the primary balance

	Year 1	Year 2	Year 10
1% increase in foreign demand for France
Real GDP	+0,27	+0,32	+0,17
Real disposable income of households	+0,13	+0,24	+0,26
Consumer price index	+0.06	+0,18	+0.58
Production price index	+0.08	+0.22	+0.66
Primary balance/GDP	+0,09	+0,16	+0,05
Oil price increase of $25/barrel
Real GDP	-0,25	-0,6	-0,55
Real disposable income of households	-0,7	-0,85	-1,35
Consumer price index	+0,8	+1	+0,6
Production price index	+0,5	+0,65	+0,05
Primary balance/GDP	-0,14	-0,5	-0,4

Source: Mésange Model.

What inflation?

Table 2 also shows that, in the event of an increase in oil prices, consumer prices rise faster than producer prices, whereas the opposite is true after a demand shock. As detailed by Fipeco (2021) and the OFCE (2022), the tax bases (income tax, social contributions, production taxes, etc.) essentially follow the GDP deflator (price of a unit of GDP), while public spending follows the consumption deflator (price of a unit of consumption). When the GDP deflator grows slower than the consumption deflator, revenues grow less than expenditures and the deficit increases. This effect can be hidden in the short term by indexation lags, but from the second year onwards it becomes clear.

Today, inflation is largely of imported origin, so the above mechanism applies. However, dynamic tax revenues are expected in 2022 due to:

Rising corporate profits in 2021, driven by a very strong recovery in the economy. The corporate tax collected in 2022 depends largely on profits in 2021.
The strong dynamism of employment in 2021: at the end of March 2022, according to INSEE, employment was 2.8% above its end-2019 level, while GDP was only 0.3% above its pre-crisis level. The greater dynamism of the wage bill than of nominal GDP yields higher revenues from tax and social security contributions.
The decline in the savings rate from an exceptionally high level in 2020 and, to a lesser extent, in 2021. The VAT base therefore grows both through a price effect and a volume effect.

It should be noted that these three effects, which are linked to the recovery from the pandemic, are temporary: the savings rate cannot decrease every year; the exceptional economic recovery of 2021 will not be repeated any time soon; the labour share could gradually return to its long-term value. Conversely, the increase in import prices could well be permanent, which will induce a volume effect (lower GDP, lower employment, lower consumption in real terms) and a value effect (lower increase in the value produced compared to the value consumed), both of which are unfavourable to the primary balance.

Debt

The speed of accumulation of public debt depends fundamentally on two factors:

The primary balance (see above)
The difference between the interest rate and the growth rate, the so-called "r-g", multiplied by the debt level.

The apparent nominal interest rate on public debt - the interest payments divided by the amount of debt the year before - was 1.5% in 2021 for France. One tenth of the debt is indexed to inflation, so that each additional percentage point of inflation leads to an immediate increase in the apparent rate of about 0.1 percentage point (and therefore a cost of €2.5bn). 30% of these securities are indexed to French inflation and 70% to Eurozone inflation. The rest of the debt is fixed rate. As the average maturity is 8.2 years (see Copin and Dalbard, 2022), the rise in interest rates following the rise in inflation takes time to be passed on in the cost of the debt (graph).

Graph : impact of a 1pp interest rate increase on government interest payments in France

Source: Finance Law for 2022, programme n°117

To make it simple, on the one hand there is an immediate effect, for the 1/10th of the debt which is indexed to inflation, and for the rest of the debt there is a progressive effect linked to the rise in market interest rates, as the debt is refinanced. Because inflation has risen much faster than interest rates so far, the former effect dominates in the short term. In the longer term, however, it is the rise in interest rates that dominates.

Such an increase in the debt burden is not a problem if nominal GDP increases by the same amount. However, the inflation to be taken into account is that of the GDP deflator, while the rise in the nominal interest rate is based on the rise in consumer prices (cf. the European Central Bank's objective). And of course, if real GDP is negatively affected by the energy shock (see Table 2), debt accumulates all the faster.

***

>> Version française : Finances publiques : une inflation qui rapporte ?

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

The ruble that hides the forest

2022-06-23T00:00:00+02:00

The effectiveness of economic sanctions is notoriously difficult to estimate because of the multiplicity of instruments, intermediate and ultimate objectives, intensities (number of countries, products, years, etc.), but also because of the difficulty of defining the counterfactual scenario (see, in particular, Trésor-Economics (2015) or, more recently, Felbermayr et al. (2020). How should the spectacular recovery of the ruble since mid-March be interpreted, after its sharp depreciation during the first weeks of the war (see Figure)?

Figure.

Budget revenues boosted hydrocarbon prices in the short term

Since the invasion of Crimea in 2014, Russia has averaged a per capita growth rate of only 0.5% per year. However, the macroeconomic situation was healthy in 2019, with public debt at only 19% of GDP, a budget surplus of 1.9% of GDP, an unemployment rate of 4.6% and a current account surplus of 3.9% of GDP, despite somewhat high inflation at 4.5% (see table, first column). In 2020, the decline in GDP was relatively limited, thanks in particular to substantial budgetary support. The budget balance deteriorated to -4% of GDP as a result of the support measures, but also of the drop in revenue linked to energy exports, which represented around 40% of budgetary resources in 2019.

Table. International Monetary Fund Spring forecasts for Russia

	2019	2020	2021	2022	2023
GDP growth (%)	2.2	-2.7	4.7	-8.5	-2.3
Consumer price inflation (%)	4.5	3.4	6.7	21.3	14.3
Unemployment rate (% of the labour force)	4.6	5.8	4.8	9.3	7.8
Government budget balance (% of GDP)	1.9	-4.0	0.7	-4.0	-5.3
Current account (% of GDP)	3.9	2.4	6.9	12.4	8.1

Source: International Monetary Fund, World economic outlook database, April 2022.

In economic terms, the first effect of the war was a tripling of the trade surplus: from January to April 2022, Russia had a cumulative trade surplus of $106.5bn, compared to $35.2bn in the first four months of 2021, according to the National Bank of Russia. This tripling is the result of the rise in hydrocarbon prices (much of which occurred before the outbreak of the war) and the collapse of imports. As a matter of facts, Western and Chinese exports to Russia fell by around 40% in value between February and March 2022. According to FourKites, Russian import volumes were down 89% in May 2022 compared to February.

With rising export revenues, falling imports and initially tight exchange controls, the recovery of the ruble is not a surprise, as Itskhoki and Mukhin (2022) have shown. In a way, Russia is accumulating too much foreign currency for its needs, as it can no longer use it to buy goods, services and financial assets abroad. The withdrawal of Western multinationals from Russia cannot be considered as a capital outflow when the counterpart is a provision for losses in the accounts of the parent companies. Itskhoki and Mukhin show that, for the Russian economy, restrictions on Russian imports have the same effect as restrictions on exports combined with a freeze on foreign assets: in both cases, imports, production, tax revenues and national income are bound to decline in the near future. But in the first case the ruble appreciates (as has been observed), while if Russian exports fall without the central bank being able to intervene in the foreign exchange market, the ruble depreciates.

In the first four months of 2022, Russia's budget revenues increased by about 34% compared to the same period in 2021, according to the National Bank of Finland's Emerging Economies Research Centre. Of the 40% of energy-related budget revenues, 4/5 come from the sale of oil. According to the same research centre, these revenues increased significantly with the price increase during 2021, and continued to increase after the start of the war, despite the discount on Russian oil - about $30-35 per barrel. However, other tax revenues fell by 20% in March and April compared to the previous 12 months. In particular, VAT revenues were halved despite inflation of over 17% year-on-year.

Expenditure is also up by 25% in the first four months of the year compared to 2021. This is due to the war itself, to government support to the economy and to households, and to inflation. The fiscal situation has started also to deteriorate as a result of the decline in oil production of around 10% in April 2022 compared to the first three months of the year.

The good performance of the ruble is not of much help when imports are constrained: the external purchasing power of the currency is largely theoretical as it is impossible to obtain goods from advanced countries.

Exports do not necessarily make you richer

Russia's trade surplus is a collateral effect of the war, not an objective of the Russian government. Nevertheless, it illustrates how misleading is the external balance to assess a country's enrichment. According to estimates by the Russian Ministry of Economic Development, GDP would have fallen by 3% between April 2021 and April 2022. The Ministry explains this decline by the pressure of sanctions, in particular logistical restrictions and the decline in domestic demand. Retail sales decreased by 9.7% over the same period.

According to the International Monetary Fund (see table above), Russia's GDP could fall by 8.5% in 2022 and by another 2.3% in 2023. The National Bank of Russia predicts negative growth of -8 to -10% in 2022, and 0 to -3% in 2023; the corresponding figures are -7.8% and -0.7% for the "central" scenario of the Ministry of Economic Development, for which the purchasing power of households could fall by almost 7% from 2022. For the record, GDP fell by 2% in 2015, following the invasion of Crimea in 2014 and subsequent Western sanctions.

So far, employment has been resilient to the downturn, partly due to temporary job retention agreements (or constraints), part-time work and other imposed leaves. However, this situation is not sustainable and companies will end up laying off workers, especially as bottlenecks in industry are expected to increase in the coming months as inventories of intermediate products run out.

The gradual deterioration of public finances, as well as the need to recapitalise some companies, particularly in the banking sector, could force the government to resort to borrowing. In the absence of access to the Western capital market, it would have to borrow essentially on the domestic market, from banks or households to whom it would have to offer an interest rate higher than inflation, with side effects on productive investment already forecast to fall by nearly 20%. The alternative solution would be to monetise the deficits, i.e. to have them financed by the central bank, with a high risk of aggravating inflation and triggering a run on bank deposits.

Russia's impoverishment could be long-lasting

Europeans will gradually move away from their dependence on Russian hydrocarbons. Asia will unlikely make for the loss of European market. More importantly, the transition to renewable energies could gradually reduce global demand and prices of hydrocarbons.

The withdrawal of Western multinationals from the Russian market and the cessation of deliveries of essential inputs to key industries such as automotive, aeronautics or IT services also signal a lasting slump in the Russian economy. According to Simola (2022a), the food import substitution strategy put in place in 2014 (in reaction to the embargo on Western exports in this sector) has not been a success: while local production has been able to substitute for imports in the area of poultry or pork, the overall effect has been a rise in consumer prices. In the field of high technology, substitution could be even more difficult. In some sectors such as electronics, transport equipment, pharmaceuticals or machine tools, the share of foreign inputs in the value produced is very high. China's share in total procurement increased by 2.5 percentage points between 2013 and 2018. However, its weight of 14% is still much lower than the combined share of "Western" countries, stable at 61%. Looking at the dynamics of Russian input imports since 2012, Simola concludes that there is limited potential for substitution, particularly in pharmaceuticals and transport equipment manufacturing.

So how can the impact of trade sanctions be assessed? Unsurprisingly, this topic is subject to the same methodological debates as for the European economies. Using international input-output tables representing value chains for 35 sectors, Simola (2022b) estimates the impact on Russia's GDP of a total halt in Russian exports and imports to/from "Western" countries (including Japan, Korea and Taiwan) at -12% and -10% respectively. In the case of complete decoupling, the impact would therefore be very significant, even if the two figures should probably not be added together insofar as the cessation of exports renders some imports useless. Using a general equilibrium model, thus taking into account substitutions and price adjustments, Felbermayr, Mahlkow and Sandkamp (2022) find that a quasi-complete decoupling would reduce the standard of living in Russia by about 10%. Langot et al (2022) obtain a similar order of magnitude using the model of Baqaee and Fahri (2021). However, these figures do not take into account the withdrawal of Western multinationals from Russia. Assuming that half of the subsidiaries' activity would be discontinued, Mahlstein et al (2022) estimate that this effect alone would decrease Russian GDP by 12%. Korhonen and Kortelainen (2022) obtain about -10% using the International Monetary Fund's integrated model, which incorporates confidence shocks, but without full trade decoupling.

From one disease to another

Over the past decades, researchers have regularly asked whether the Russian economy could be subject to the “Dutch disease.” This syndrome describes the difficulty of energy and commodity exporting countries to diversify their economies by developing manufacturing industry in particular. This is because the oil rent attracts the factors of production - labour and capital - and increases the costs for other activities which then struggle to emerge in the face of foreign competition. Only those sectors that are "sheltered" from foreign competition develop, mainly in services. The phenomenon can be reinforced if the government distributes the tax revenues linked to oil rents and, on the contrary, mitigated if, as is the case in Norway, the government accumulates the corresponding revenues in a sovereign fund instead of spending them. From 2008 onwards, Russia has saved part of its oil revenues in a fund, following different fiscal rules. However, the assets of the fund seem to have been severely devalued since the beginning of the war, not to mention the sanctions on the part invested abroad.

The programmed attrition of the energy rent will keep Russia away from the Dutch disease, even if it will leave it with other raw materials that will certainly be able to increase in value in the coming period: nickel, palladium, cereals... However, Russia will have to develop new economic activities and insert them into the global value chains. This will involve bringing back foreign investors rather than relying on import substitution, which does not seem to have worked well in the 2014-2021 period. Contrary to what the reputation of the Russian mathematics school might suggest, the education system does not seem to be performing particularly well when assessed by the PISA tests at age 15; Russia is ranked only 45th in the Global Innovation Index 2021, between India and Vietnam. It will be difficult to find a path to growth without reopening the country, attracting capital, technology, know-how and talent, especially as Russia's population could decline by 0.3% per year by 2030, according to the UN median scenario. The Covid pandemic and the war could accelerate the demographic decline: in 2021, the country lost more than one million inhabitants (about 0.7% of its population) due to more deaths than births, in relation with the pandemic; and even though it benefited from a positive net migratory inflow, Russia's total population fell by about 0.4% that year. The war could dry up or even reverse migrations.

***

>> Version française : Le rouble qui cache la forêt

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Energy crisis: Europe by candlelight?

2022-05-27T00:00:00+02:00

The International Monetary Fund has revised downwards its forecast for world growth for 2022 (-0.8 percentage points) and for 2023 (-0.2 pp). Unsurprisingly, Europe is heavily affected by Russia's invasion of Ukraine and its geopolitical consequences, to varying degrees depending on the energy mix, exposure to Russia and the share of industry in each country's economy (Figure 1).

Figure 1. Revised IMF growth forecasts for 2022 and 2023

(deviation from the January 2022 forecast, percentage points)

Source: IMF, World Economic Outlook, April 2022.

Of course, the revision of the growth outlook is not only due to the war in Ukraine. The current epidemic wave in China is also clouding the forecast. However, the war is a major factor in this revision, taking into account the situation as of 31 March 2022, with oil prices of nearly US$107 per barrel in 2022 and US$93 in 2023 (compared to an average of US$91 at the beginning of the year).

Numerous figures on the impact of the war on the European economy have been produced over the past two months, sometimes with different orders of magnitude. How can we make sense of them? We can proceed in two steps: (1) which shocks? (2) which transmission channels to the European economies?

Which shocks?

Although it only represents 1.75% of world GDP (2020 figure in current dollars, source: World Bank), Russia is a major world player in the production and export of hydrocarbons, but also of fertilisers, metals, cereals and wood. The war is severely affecting these supplies.

For European economies heavily dependent on Russian hydrocarbons, the energy shock will probably be the most painful. To calibrate it, the question of oil (world market) and gas (local market) must be treated separately.

Crude oil

Even before the war began, crude oil prices rose sharply as a result of the economic recovery from the Covid crisis, while supply had not fully returned to its pre-crisis level (Graph 2). The war has reinforced this movement, as Russian supply has decreased due to transport and payment difficulties in connection with the sanctions. According to the International Energy Agency, deliveries may have already fallen by around 2.5 million barrel-days (or 3% of global supply). In the event of a European embargo on Russian crude oil, 3.4 million barrel-days would disappear from the market, i.e. 4% of world supply, if we disregard re-directions to other countries, which would probably be limited despite the discount they enjoy.

Figure 2. Daily crude oil price (Brent, current US$ per barrel)

Source: Fred, Federal Bank of St Louis. The period of war is in red.

Between the period January-February 2022 (up to and including 23 February) and the period 24 February-25 April, the price of crude oil increased by 22% (from $91 to $111 per barrel). A strong price response (+22%) to a moderate decline in expected global supply (-3%) means that global demand and non-Russian supply are not very price sensitive (or elastic): the price has to rise a lot to convince oil producers to increase their supply and to discourage some demand, and thus clear the market.

The sum of their elasticities is here implicitly equal to 3%/22% = 0.14 over the period. By applying these same elasticities to a decline in world supply of 4% rather than 3%, the price increase would be 4%/0.14 = +29% compared to the pre-war level, which would put the price of a barrel at 117 dollars.

Using short-term elasticities more in line with the academic literature (of the order of 0.1 for both supply and demand, cf. Labandeira et al., 2017, Caldara et al. 2019), the price increase would be 4%/(0.1 + 0.1)=+20%, which would put the price of a barrel at 109 dollars. Conversely, the demand subsidy implemented in many countries could weaken price elasticities. In the end, if the price elasticity of demand were zero, i.e. if demand were totally rigid whatever the price before subsidies, the same simple model would lead to a price increase of 4%/0.1=+40% compared to before the war, and thus to a barrel of 127 dollars.

Some experts instead estimate the additional production margins in the rest of the world, and then ask how much the oil price must rise to bring demand down to the level of supply. From an economic rationality point of view, this approach is equivalent, as prices have to rise to convince suppliers to extract more oil. However, this alternative approach better takes into account the fact that oil supply is actually held by a cartel - OPEC - which has so far shown no sign of deviating from its initial production trajectory, despite rising prices and unused margins. This is a bullish factor for prices.

Conversely, a bearish factor is often overlooked: the decline in global activity. Assuming that world GDP is reduced by around 1% (IMF figure, see graph 1), world demand for crude oil could be lower by around 1%, reducing the shortage and therefore the price increase. This effect is included in the macroeconomic models simulating the impact of the war on the world economy, but not in those that consider a country in isolation, the price of oil being an assumption outside the model.

As can be seen, there is therefore a great deal of uncertainty about the evolution of the price of crude oil, but this is nothing compared to gas.

Natural gas

The case of gas differs from that of oil because European imports are mainly delivered via gas pipelines: due to transport constraints, the gas market is not global but rather regional, and prices are not unified (graph 3). Gas prices have risen sharply in Europe in 2021, as the sharp increase in demand linked to the economic rebound has been met with less dynamic supply from the Netherlands and Russia, the latter having gradually stopped serving short-term markets (honouring only its already signed long-term contracts).

Figure 3. Monthly natural gas prices, in US dollars per mmbtu*.

Sources: World Bank and IEA. * Million Metric British Thermal Unit: 20 mmbtu are equivalent
to about 6 MWh. Last point : April 2022.

In 2020, Russia produced 22% of the world's gas while Europe consumed 13% of the natural gas produced in the world (Enerdata figures). As Russia accounts for around 40% of the gas consumed in Europe, the disappearance of this supply would represent a shock of 13%x40%=5% at world level but much more at regional level, as alternative supply is limited by the transport and re-gasification capacities of imported liquefied gas (LNG), which today accounts for only 20% of European supply (source Bruegel). Replacing Russian gas imports entirely with LNG would mean tripling European LNG supplies, which in the short term is neither technically possible (limited supply on the world market, low additional regasification capacity in Europe) nor economically feasible (Europe is in competition with Asia on the LNG market, and redirecting flows to Europe is costly). In this respect, the IEA envisages a substitution of only 13% of the missing Russian gas.

This is why, since 2021, the price of gas has risen much more than the price of oil for Europeans: around +60% in March 2022 compared to February; and x5 in April 2022 compared to April 2021. And this is why, in the event of a disruption of Russian gas supplies, experts envisage an extremely high price, or even quantitative rationing of gas on European soil.

Other shocks

In addition to crude oil and natural gas, Russia exports coal (which is transported by ship, making it highly substitutable with other suppliers) and refined oil products, particularly diesel, for which Europe is particularly dependent, as refining capacities are specific and difficult to substitute in the short term. Russia also exports rare metals (nickel, palladium, of which it is the world's leading producer) for which substitution is delicate, and other products (fertilisers, wheat, wood, etc.) which are traded on a world market.

In addition to these supply and price shocks, Europeans have to add the loss of markets in Russia due to export restrictions, which in May will account for around 60% of the volumes exported to that country in 2021, or around 0.4 percent of European GDP. Moreover, the withdrawal of large European companies from Russian territory would represent around 1 percent of GDP in losses for the European economy according to Blanchard and Pisani-Ferry (2022).

Finally, the invasion of Ukraine could trigger precautionary behaviour on the part of households and companies in Europe, leading them to revise their investments downwards, to save more and to direct their savings towards low-risk assets. It could also increase uncertainty in financial markets. However, the depreciation of the euro (-8% against the US dollar between 23 February and 12 May 2022) could support activity in the euro area in the short term, while increasing inflationary pressures.

Transmission channels to the economy

Once the shocks have been defined, we need to look at the transmission channels to the European economies. Here we focus on the energy shock.

Crude oil

Oil price volatility is not new. In real terms, the price increase observed at the end of 2021 and the beginning of 2022 is not exceptional (Graph 4), and modellers are used to simulating the impact of such a shock.

Figure 4: Real crude oil and gas prices

(monthly data, base 100 = average 1960-2021)

Sources: Commodity Research Bureau and Bureau of Labor Statistics.
The crude oil price is deflated by the US consumer price index. Last point : March 2022.

For a net importing country, a rise in oil prices leads to a transfer of income to the rest of the world, and thus to impoverishment. If wages do not immediately adjust to the price increase, purchasing power and thus consumption will decline in the short term. Companies, for their part, cannot immediately pass on the higher energy prices in their sales prices, so their margins shrink, to the detriment of investment. When companies raise their prices, they preserve their margins but lose market share.

However, the rise in oil prices is a global shock. As all companies face higher costs, the competitiveness effects are limited. On the other hand, falling incomes in other oil-importing countries mechanically reduce foreign demand, and thus exports. In total, an increase in oil prices of 25 dollars reduces activity by about ¼ point over 1 year in a country like France (table).

Impact on GDP of a $25/barrel oil price increase in a neo-Keynesian model

(in % deviation from a scenario without shock)

Model	Country/area	International linkages	1 year impact
Oxford Economics	Euro zone	yes	-0,25
E-Mod (OFCE)	France	no	-0,25
ThreeME (OFCE)	France	no	-0,29
Mésange "naïve"	France	no	-0,25
Mésange "Reamistic"	France	yes	-0,3

Source : model documentation, neglecting possible non-linearities.

Natural gas

In standard macro-econometric models, gas is not identified as such and its price is assumed to follow that of oil. A fortiori, a gas supply disruption is difficult to simulate. In these models, the production of goods and services depends on labour, capital and an exogenous total factor productivity (TFP). A disruption in the supply of imported gas could be seen as an exogenous decrease in TFP. However, in these Keynesian models, TFP is only relevant in the long run. In the short run, GDP is determined by demand, although prices respond to supply shocks.

To properly account for the disruption of value chains, one must instead use a general equilibrium model describing how the shock affects not only industries using gas directly, but downstream industries - chemicals, glass, etc.; and how it can be cushioned by substitutions and the use of imports at all levels of the value chains. Said differently, German companies can choose to import energy-intensive goods, or seek less energy-intensive substitutes. Using the Baqaee and Farhi, 2021 model, Bachmann et al (2022) calculate the effect on the German economy of a 10% drop in energy supplies (gas not being isolated but grouped with other energies), according to two methods: a simplified formula, which leads to a loss of GDP of about -1%; and a full model with international linkages, according to which the same shock reduces German GDP by only -0.3%. Due to a lower dependence on Russian gas, the same calculations for the French economy lead to more limited GDP losses, between -0.15% (full model) and -0.2% (simplified model), see Baqaee et al. (2022).

In these general equilibrium calculations, prices and wages are assumed to adjust perfectly - we are talking only about relative prices. To "compensate" for this shortcoming, the authors assume a relatively rigid production: substitutions between intermediate goods are difficult. However, in a rigid economy, the rebalancing of markets after a shock implies strong variations in relative prices, and therefore a significant economic cost when certain prices only adjust with delay. For example, companies in the glass sector fail to pass on higher costs to their customers, and some stop production.

Therefore, a combination of different approaches is needed to arrive at a realistic estimate of the effects of the energy crisis. Other elements of the crisis can then be added - falling exports to Russia, depreciation of certain assets, precautionary behaviour, public policies, etc.

Orders of magnitude

Having identified the main methodological knots, we can now better understand the differences between the various figures that have been circulated, a sample of which is summarised here. The economic cost of the crisis is all the greater when (i) the energy price hypothesis is high, (ii) macro-Keynesian effects are taken into account, and (iii) the productive fabric is rigid.

Blanchard and Pisani-Ferry (2022) estimate the drain on European incomes due to a 25% increase in the price of imported oil and gas at around 1 point of GDP. This figure is consistent with the IMF's revised growth forecast for 2022 (Graph 1). It constitutes a form of hard core economic cost of the war to date for Europeans - a cost that could increase as a result of higher prices, supply disruptions, unfavourable international spillovers, wealth effects or more cautious behaviour on the part of households and businesses. If the energy crisis is temporary, governments will be able to mitigate its effects. However, if the crisis lasts, the loss will have to be progressively distributed among the various economic agents.

***

>> Version française : Crise énergétique : vers une Europe à la chandelle

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

European fiscal rules in the air

2022-02-28T00:00:00+01:00

A long journey

To think about a reform of European fiscal rules, it may be useful to go back to 1992, when the Maastricht treaty was signed. At that time, macroeconomic policy-making was shaped by the nascent New-Keynesian theory that clearly separated the long term (the “steady-state”) from the short term, with monetary and fiscal policies only having a temporary impact on the real economy. Furthermore, following up on Barro and Gordon (1983), it was believed that central bank independence would fix the time-inconsistency problem of monetary policy: separating monetary from fiscal policy, and assigning a single objective (price stability) to an independent central bank would eliminate the inflation bias, at no cost in terms of employment in the medium run. Incidentally, stabilising inflation would indirectly contribute to output stabilisation in the event of demand shocks, which were considered dominant after the disinflation period of the 1980s was over.

In this context, the European Central Bank (ECB) was assigned euro-wide price stability as its main objective, whereas national governments would use their fiscal policies to react to country-specific shocks. To ensure they had enough fiscal space to do so, and also due to the impact of a sovereign debt crisis on euro-wide financial stability (given the interconnections within the financial sector, see Eichengreen and Wyplosz, 1998), governments were asked to keep their deficits and debts below 3% and 60% of GDP, respectively. They also committed to reach fiscal balance over the medium term. In brief, the organisation of monetary and fiscal policies was considering price stability and fiscal sustainability as the two “common goods” that had to be preserved within a monetary union (see Table 1, 1^st column).

Table 1. Monetary and fiscal policies in the euro area: moving away from Maastricht

	Maastricht	Post global financial crisis	Post Covid
Policy assignment*	M for common shocks F for country-specific shocks	M also for country-specific shocks (OMTs) F also for common shocks (ZLB)	M, F also having to contribute to a smooth green transition
Shocks	Mostly demand shocks	Also financial crises, sudden stops	Tail risks, supply shocks
Common goods	Price stability, fiscal sustainability	Also financial stability, safe assets	Also climate

*M stands for monetary policy; F stands for fiscal policy.

After the 2008 global financial crisis and subsequent euro area crisis, the assignment of monetary and fiscal policies started to depart from the Maastricht clear separation (Table 1, 2^nd column). One reason was the introduction of the “whatever it takes” instrument, namely the outright monetary transactions (OMTs) whereby the ECB could buy potentially unlimited amounts of a specific country’s sovereign debts on the secondary market, conditional on an adjustment programme, in order to “safeguard an appropriate monetary policy transmission and the singleness of the monetary policy” (ECB, 6 Sep. 2012). To achieve this objective, the ECB was recognising its role to buffer idiosyncratic market shocks when the integrity of the euro area is at stake.

A second reason for departing from strict policy separation was the zero lower bound constraint on policy rates and subsequent re-discovery of the policy-mix: fiscal policy had to step in not only in case of idiosyncratic shocks, but also as a complement to monetary policy (see Draghi, 2014). However, the need for such aggregate fiscal stance remained controversial, and it took time before the pro-cyclicality of aggregate fiscal policy in 2011-2013 was recognised.

With the Covid crisis, in 2020, monetary and fiscal policies became fully congruent at euro area level, with the combination of gigantic fiscal support at national and European levels (cf. NextGenEU package and SURE instrument), and of the Pandemic Emergency Purchase Programme (PEPP) by the ECB, on top of its Asset Purchase Programme (APP). The pandemic also acted as a wake-up call concerning climate change, and the policy assignment decided in Maastricht was further rethought: shouldn’t monetary policy contribute, e.g. through its collateral policy, to the needed reallocation of private investments in favour of climate change mitigation? Shouldn’t fiscal policy secure public, green investments over a long period, at the expense of other spending and/or debt reduction? In a nutshell, joint European commitments to reduce greenhouse gas emissions had to be considered as another common good, and macroeconomic policies could not ignore this fact (Table 1, 3^rd column).

The strong rebound of the economy in 2021 revealed weaknesses in global supply chains and in energy provision. The steep increase in energy prices, progressively passing on manufacturing and consumer price indices, sounded as a reminder that the economy could be hit not just by demand shocks, but also by supply shocks. Furthermore, the energy transition could well revive such cost-push shocks in the future.

Last but not least, European economies are recovering from the crisis with new legacy debts, and debt-to-GDP ratios have sometimes reached levels that are far from the Maastricht, 60% benchmark.

Which landing area?

It is tempting to assign a long list of objectives to fiscal policy. Despite the multiplication of challenges, though, the objective of fiscal rules remains to ensure fiscal sustainability and raise the predictability of fiscal policy. Adding other objectives would weaken the ability of fiscal rules to reach any of them.

That said, fiscal rules need to adapt to the landscape of high debts and low real interest rates. They also need to enable or even incentivize counter-cyclical policies both in economic downturns and upturns. Finally, they need to be simpler, better enforced and less dependent on unobserved variables (notably the output gap), while at the same time leaving room for judgement and taking into account country-specific situations – somewhat conflicting requirements.

The way to reform fiscal governance in the euro area has been hotly debated among experts since the introduction of the Stability and growth pact, in 1997, and after each reform (mainly 2005 and 2011). After the Covid crisis, some convergence is now visible around four lines (see, e.g., “7+7” report, 2020; European Fiscal Board, 2020; Martin, Pisani-Ferry and Ragot, 2021):

The need to rethink the debt objective: a gradual decline of the debt-to-GDP ratio is warranted to rebuild fiscal buffers and protect governments from self-fulfilling crises as well as large exogenous events. Depending on long-term growth, the same deficit can lead to different debt dynamics, hence to different assessments in terms of fiscal sustainability. This explains the need to focus on debts rather than deficits. However the debt ratio is only partially in the hands of the government, hence the need to complement the debt path with a more operational rule.
An expenditure rule in place of structural balance: in order to fulfil the debt adjustment target, capping expenditure growth seems more appropriate than monitoring the reduction of the structural deficit, since (i) it is easier to communicate and less prone to measurement errors, (ii) it leaves more room for counter-cyclical fiscal policy during a downturn, and (iii) it enforces counter-cyclical fiscal policy also when the economy is booming (see e.g. IMF, 2018).
A differentiation across countries: the 60% debt threshold, dates back to a time where nominal growth was believed to be around 5% yearly on average (hence debt could be stable at 60% of GDP with a 3% deficit). This uniform debt benchmark is now widely recognised to be inappropriate as a medium-term guideline: potential growth is down, debt levels are up, and the never-ending world demand for “safe assets” suggests that higher debt levels can meet a stable demand with still low real borrowing costs. The 60% threshold is especially problematic when combined with the “1/20^th rule”, which requires that 1/20^th of “excess debt” be eliminated each year. This combination could end up in excessive, pro-cyclical and/or unrealistic fiscal tightening in some countries. In fact, different positions in the cycle and different growth perspectives require different deleveraging speeds.
Compliance and ownership: although the 3% threshold seems to have exerted a “magnet effect” (see Caselli and Wingender, 2018), compliance with the debt rule and medium-term objectives has been somewhat disappointing (European Fiscal Board, 2019). One reason has been the lack of ownership at national level, which itself can be explained by (i) the complexity of the rules, (ii) the limited credibility of the sanctions, and (iii) the “one size fits all” feature of the rules. Simplifying the rules, turning sanctions into incentives, and giving more leeway to governments to propose a debt trajectory are different ways to raise national ownership.

Safeguarding investment

Previous experience of fiscal adjustment in the euro area has demonstrated that sharp spending cuts frequently fall on public investment disproportionally. In reaction, an “investment clause” was introduced in 2015 as a flexibility of the Stability and growth pact, but with very strict strings attached. To start with, only those investments that are co-financed by the EU are eligible. Furthermore, the clause can be called only when the output gap is below -1.5%. Finally, only limited and short-lived deviations of the structural deficit are allowed (less than 0.5 pp, the financial deficit having to stay below the 3% boundary). In brief, the existing investment clause seems to be of little help for the problem at stake, which is to heavily co-invest (along with the private sector) in the green transformation of the economy.

In order to safeguard investment, two routes have been proposed:

A (green) golden rule (Darvas and Wolff, 2021): excluding “green” investments from the calculation of expenditure (or deficit) ceilings would encourage governments to cut current expenditure rather than public investment expenditures. Aside from the risk of “greenwashing”, the main objection to such proposal is that it would lead to departing from the prime objective of the fiscal rules, which is debt sustainability. Granted, accumulated GHG emissions are also a liability and unmitigated climate change would undoubtedly lead to large damage costs and lower GDP, and hence also be a threat to debt sustainability. However, it remains that national debts will have to be served with national budget resources and continuous debt rollover, regardless of the type of expenditures that have been financed through debt. Another problem is that excluding a category of spending would alleviate the pressure to make these expenditures efficient (as measured for instance by their abatement costs) and introduce inefficient distortions between different kinds of spending (e.g. public funding for housing retrofitting would be considered as a green investment but not professional training in this area). Finally, a (green) golden rule would depart from the necessity to simplify the rules and it would require a common and precise definition of what constitutes a “green” investment.
A climate investment fund (Garicano, 2022): since green investment is mostly targeted to achieve a common commitment concerning greenhouse gas emissions (Paris agreement, Fit-for-55 package), a common funding of related investments makes sense. Such an approach would allow to prioritise investment projects across Europe based on their abatement costs, using a single methodology, and building on the experience of NextGenEU. Such a fund could be transitory, ending for instance in 2050 in line with the horizon of EU emissions commitments. It would require a new increase in own resources, which in turn would raise the question of the transfers across the member states. Absent such transfers, a climate investment fund would be close de facto to a golden rule, each Member state financing its own investment (see Darvas, 2022). However, it would allow for financing a key economic and political ambition with a common fiscal tool, safeguarding the simplicity of fiscal rules and incentivising all Member states to invest more in the green transition. Another possibility would be that those countries with lower abatement costs achieve more ambitious reductions in GHG emissions and receive more funding for this purpose. Last but not least, the access to the fund could be conditioned to compliance with fiscal rules, turning existing sanctions into incentives.

The debate on how to safeguard investment will not be settled overnight. The problem is that the commitments to invest in climate change mitigation and adaptation need to be taken without delay. How to square the circle? One possibility could be to include “green” reforms and investments (both physical and human) as one element that would modulate country-specific fiscal requirements upstream, together with the debt level, foreseeable growth prospects, and existing non-fiscal imbalances (the latter should not be forgotten when assessing fiscal sustainability, see e.g. Koll and Watt, 2022). Each Member state would propose a comprehensive strategy to the Commission that would include three elements: (i) a fiscal adjustment path (debt anchor and expenditure growth limit); (ii) concrete green investments and reforms; and (iii) a credible response to macroeconomic imbalances as identified in the country-specific recommendations. After agreement with the Commission, this comprehensive strategy would be considered as a hard commitment from the Member state. The translation of the process across multiyear and yearly commitments would still need to be worked out. Such an approach would be more holistic than the golden rule approach. For instance, a country experiencing tensions in its construction sector and a current-account deficit could be asked to reduce public subsidisation of house retrofitting and accelerate fiscal adjustment accordingly. Conversely, a country with excess savings could be asked to accelerate its investment programme, even though this may temporarily increase its government deficit. Such an approach would be compatible with the introduction of a common investment capacity if that capacity was to be agreed on, but the latter would not be a precondition.

Since monetary unification, much has been learnt about the functioning and deficiencies of the initial setup. Concerning fiscal rules, there has been a tendency to progressively add layers of complexity. The challenge of decarbonisation should not be another occasion to add a layer. Rather, it could be used as an opportunity to adopt a broad approach to sustainability, encompassing its fiscal, financial, macroeconomic and environmental dimensions, and providing the right incentives at national level. Raising ownership upstream could help stricter enforcement downstream.

***

>> Version française : Règles budgétaires européennes : comment atterrir ?

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Ex ante, ex post: tuning the two pillars of policy evaluation

2022-02-25T00:00:00+01:00

My last post, on the simulation and analysis of firm-level data to inform policy-making during the crisis, did not give rise to any complaint. Still, it was describing how micro-simulations carried out at the beginning of the crisis, which anticipated a sharp increase in insolvencies, have been partly contradicted by the actual financial statements, which so far have evolved more favourably. Did the modellers underestimate the capacity of the economy to adapt? Or are the financial data for 2020 misleading? In short, should we give more credit to ex ante micro-simulations or to ex post accounting observation?

Even though year 2020 was marked by the pandemic and the emergency support granted to companies and households, we should not forget that technical progress, climatic hazards and changes in consumers’ preferences and behaviours were not put on hold. Thus, the simple comparison of the 2019 and 2020 financial statements of companies does not allow to conclude on the impact of the crisis nor on the impact of public support: it is an accounting observation rather than a genuine ex-post evaluation, which would require comparison with a counterfactual scenario without crisis and/or without public support.

Evaluation of the recovery plan

The evaluation of the recovery plan will lead to the same challenges. When it was designed, macroeconomic simulations using the French Treasury's neo-Keynesian model, Mésange, estimated its cumulative impact on GDP over the period 2020-2025 at +4%, excluding microeconomic effects and European spillovers (Figure 1). A Recovery Plan Evaluation Committee was then commissioned to evaluate the effective ex-post impact. This committee has just launched a series of calls for proposals to evaluate some flagship measures such as energy retrofitting of public and private buildings, the '1 young person = 1 solution' plan, etc. - but also to evaluate the plan as a whole. Considering the different measures altogether, and accounting for their interactions, what will have been the impact of the plan on GDP and employment compared to the natural rebound of the economy after the crisis?

Figure 1. Estimated impact of the recovery plan on GDP

(Difference in percent compared to a scenario without a recovery plan)

Source : Rapport Economique, Social et Financier 2021, October 2020. DG Trésor, calculations with the Mésange model.

Legend: yellow stands for company taxation; orange for skills and employment; red for innovation; purple for public investment; blue for other public spending; grey for household support; shaded blue for spillover effect; shaded red for other measures not directly assessed.

Interpretation: In 2022, the total impact of the recovery plan (compared to a scenario without the recovery plan), is evaluated at +1% on GDP.

N.B: The spillover effects and the effects of other measures not directly assessed are only presented for 2021 but will continue in subsequent years.

Ex post evaluation is unfortunately one of the areas where macroeconomics has the most room for improvement. The major difficulty is the counterfactual scenario, when the whole country is 'treated' by the recovery plan, rather than a sub-group of individuals or firms. One possibility is to compare the economic performance of the 'treated' country to that of one or more 'untreated' but similar countries. For example, Campos, Coricelli and Moretti (2019) estimate the impact of European integration on GDP per capita to be around +10% over a 10-year horizon. This method is ineffective for the recovery plan because all advanced countries have deployed similar plans simultaneously, all the more so in the European Union where the plans have been coordinated. Even if a control group could be identified abroad, it would still be difficult to disentangle the effect of the recovery plan from the effect of the support measures that preceded it (with an overlap in 2021) and the effect of the France 2030 investment plan that follows it, also with an overlap.

Pension reform

To illustrate these methodological difficulties, another example is a pension reform aiming at raising the retirement age (age of entitlement and/or duration of contribution required for eligibility to a full pension). A first approach consists in simulating a macroeconomic model in order to evaluate ex ante the impact of such a reform. In a standard neo-Keynesian model, an exogenous increase in the labour force (by raising the retirement age) leads to a short-term imbalance in the labour market, which is progressively resolved by a decline in real wages. Depending on the calibration of the model and on the assumptions made (notably on the replacement income for unemployed or inactive seniors compared to retirement pensions), the increase in employment and GDP may take a while. GDP may even fall in the short and medium term according to the simulations carried out by the OFCE for the High Council on the Financing of Social Protection (Haut conseil du financement de la protection sociale, HCFiPS), while the impact is positive but limited in the simulations carried out in 2016 by the French Treasury for the Pensions Advisory Council (Conseil d'orientation des retraites, COR).

However, one may question whether a neo-Keynesian model, designed to study demand-side policies, is adequate to assess the impact of supply-side measures such as an increase in the labour force. Some will prefer to rely on an overlapping generation model such as the International Monetary Fund's GIMF model, which allows for inter-temporal effects. In this type of model, an increase in the retirement age raises the level of GDP very quickly, without any short-term depressive effect. One reason is that a fraction of households anticipate that they will be able to work longer and thus earn more money. As a consequence, they reduce their savings rate, which increases consumption. However, such simulation remains an ex ante exercise, with no guarantee of being realistic given the many assumptions on which they are based.

A completely different approach consists in analysing the impact of a similar reform such as the 2010 reform, which raised the age of entitlement from 60 for the 1950 generation to 62 for the 1955 generation. Dubois and Koubi (2017) and Rabaté and Rochut (2019) have studied the outcome for the generations affected by the reform compared to those immediately preceding them, over a relatively short time horizon. Their results indicate that the people affected by the reform have, on the whole, remained in the state they were in - employment, unemployment or inactivity. Mechanically, the reform increased the number of unemployed since some workers who would have been retired before the reform stayed longer in the labour market. However, the effect does not go beyond this mechanical shift, as shown in Figure 2: unemployment rates just follow the retirement age. In Figure 3, the employment rates increase at all ages but more specifically for the 60-61 year olds.

Figure 2. Share of 1950 and 1952 generations unemployed as they get older

On the left, the figure shows the share of unemployed women depending of their age. On the right, the figure shows the same chart for men.

Interpretation: Among women born in 1950, 10% were unemployed at 59 years and 9 months but only 4% at 60 years and 3 months. For women born in 1952, the proportions are 11% and 10% respectively.

Source: Dubois and Koubi (2017).

Figure 3. Share of 1950 and 1952 generations in employment as they get older

On the left, the figure shows the share of women in employment depending of their age. On the right, the figure shows the same chart for men.

Interpretation: Among women born in 1950, 33% were in employment at 59 years and 9 months and only 21% at 60 years and 3 months. For women born in 1952, the proportions are 42% and 37% respectively.

Source: Dubois and Koubi (2017).

This type of analysis does not take into account the possible substitution between workers of different age groups, and in particular between older and younger workers. This issue has been examined in different countries. Boeri, Garibaldi and Moen (2021) provide a recent estimate on Italian data, suggesting a limited substitution and rather with intermediate age groups. However, there does not seem to be a consensus on these substitution effects (see the review by Gruber, Milligan and Wise, 2009). When we take into account the reallocation of workers between firms, and therefore the labour market as a whole, we do not find much. Thus, Ben Salem, Blanchet, Bozio and Roger (2010) do not find any robust impact of the policies aiming at the withdrawal of senior workers from the labour market implemented before the 2000s on the professional integration of young people in France. For their part, Hairault, Langot and Sopraseuth (2009) highlight a positive effect of the increase in the retirement age on the employment rates of slightly younger workers, via a distance-to-retirement effect.

'Accounting' approach

Building on these various studies, we can extrapolate the effects of the 2010 reform to a possible further increase in the age of entitlement to retirement, say by two years at a rate of one quarter per generation, using an 'accounting' approach. To do this, we observe the distribution by status - employment, unemployment, inactivity - of people aged 62 in 2019; then we apply these proportions to the future population aged 63 or 64, potentially affected by such a reform. This calculation leads to an estimate of 240,000 (after 5 years) and 390,000 (after 10 years) additional persons in employment. Assuming that capital adjusts, this 1.4% increase in employment at the 10 year horizon results in an additional GDP of 1.4% and, consequently, an increase in fiscal revenues of around 0.6 percent of GDP, which come on top of the 0.5 pp improvement in the pension balance that results directly from the reform at this horizon. We must then subtract the additional expenditure linked to the probability of 63-64 year olds being unemployed or inactive. According to Dares and Drees (statistics departments of the Labour and Health ministries), this cost is around 0.2 percent of GDP. This results in a net gain of almost one percent of GDP for public finances after 10 years (Table 1).

Table 1. Impact on the budget balance of a 2-year increase in the age of entitlement to pension (3 months per generation), 'accounting' approach

This approach ignores possible side effects on other categories of workers for which, as mentioned before, there is little solid evidence. It also assumes that the decrease of recruitments in constrained sectors (notably the public sector) is compensated by more dynamic hiring in other sectors. This assumption is reasonable in the light of the experience of the 2010 reform, which was nevertheless concomitant to a long economic crisis. It is even more reasonable in the view of current and future labour shortages, particularly in the key sectors of ecological, digital and demographic transitions. It nevertheless assumes a progressive adaptation of skills to the jobs offered.

While it is impossible to fully dissociate the effect on employment of the 2010 pension reform from that of the global and European crises, the mixed short-term impact of an age reform, highlighted by ex ante neo-Keynesian simulations, does not seem to be verified ex post. Perhaps these models underestimate the ability of the labour market to absorb additional senior workforce, for instance because they do not distinguish between different age groups, or because they underestimate the speed of adjustment of the productive capital stock. Or else, these models are estimated over a somewhat earlier period when the economy was less flexible. Or, maybe these models do not take into account any expectation effect that may be favourable to consumption and activity in the short term. In any case, the experience of the 2010 reform is rather reassuring regarding the capacity of the French economy to absorb a positive shock to the labour force, especially in a period of labour shortage.

***

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Firm-level data throughout the Covid crisis in France

2022-01-17T00:00:00+01:00

Over the last decade, a rising amount of French firm-level data has been made available to researchers. Anonymized financial statements and tax records can be recovered through a Secure data hub, upon authorization granted by the Statistical confidentiality committee. Throughout the crisis, the French Ministry for the Economy, Finance and Recovery has used available data intensively as soon as they were made available. Three successive stages may be distinguished.

Stage 1 (March-August 2020): ex-ante analysis

At the beginning of the Covid crisis, in March 2020, the need to protect the corporate sector from the damage of the lockdown emerged very quickly. That said, the design and calibration of such support was not straightforward. At that time, only financial statements for 2017 were available in a consistent format. In order to quantify financial needs, microsimulations were run by the Treasury. Specifically, monthly sector-level shocks provided by Insee (17 sectors) for 2020 were applied to firm-level turnover for 2017 (recovered from financial statements). Fixed costs were assumed to be… fixed, and some assumptions were made on firms’ ability to adjust their variable costs over time. Then, depending on the eligibility criteria and calibration of the support schemes (mainly short-time work and subsidies through the solidarity fund), firm-level cash-flows were simulated on a monthly basis. Assuming no credit constraint (thanks to the quasi-automatic distribution of state-guaranteed loans by the banking sector), a firm with negative cash-flow would accumulate debts, on top of existing debts measured at end-2017. When debts would exceed total assets, the firm would be deemed insolvent.

Other institutions ran similar types of microsimulations, sometimes using the international but non-exhaustive database Orbis which we do not consider here. In July 2020, Guerini et al. (2020) were the first to publish a cumulated share of newly insolvent firms due to the crisis (Figure 1).

Figure 1. Cumulated share of newly insolvent firms, Mar 2020-Mar 2021
(% of initially solvent firms)

Note: simulations based on end-2017 financial statements and on a fall in activity by an average -32% during 2.5 months compared to pre-crisis, followed by gradual normalization up to -5% in September 2020 and -2% at end-2020. The solidarity fund is not accounted for.

Source: Guerini et al. (2020).

Stage 2 (Sep. 2020-June 2021): real-time data

There were two major problems with the first vintage of microsimulations. First, they were relying on sector-level activity shocks, while it soon became clear that intra-industry and geographic heterogeneity were major features of the crisis (see Barrot, 2021). Second, assumptions had to be made about the adjustment of payrolls and about the take-ups of support schemes.

Starting in September 2020, these two pitfalls could be fixed by using almost real-time data: (i) individual VAT (value-added tax) returns, providing proxies for firm-level activity shocks (except for very small firms), and (ii) administrative records of firm-level payrolls and take-ups of support schemes. Subsequently, a second wave of microsimulations was performed by Hadjibeyli, Roulleau and Bauer (2021a) at the Treasury, based on end-2018 financial statements that had become available in the meantime. The results are summarized in Figure 2: without a crisis, around 3.6% of initially solvent firms would have turned insolvent between March and December 2020. The Covid crisis would raise this proportion to 11.9% before government support. Short-time work would reduce it by 2.9 percentage points, the solidarity fund by an additional 2.1 pp, and tax deferrals and exemptions by 0.3 pp. Accounting for the three support schemes, the proportion of newly insolvent firms would be “only” 6.6% by Dec. 2020 (note that insolvent firms do not necessarily default).

Figure 2. Cumulated shares of insolvent firms over Mar-Dec 2020
(in % of initially solvent firms)

Source: Hadjibeyli, Roulleau and Bauer (2021a).

Similar research was carried out by Banque de France and Insee, with additional assumptions concerning intercompany credit, fixed capital investment and the distribution of dividends, but with less detail on government support (see Bureau et al., 2021). Hadjibeyli, Roulleau and Bauer (2021b) have compared the results from a method based on observed payrolls and support take-ups, to those obtained without this information, hence “fully simulated” results. As evidenced in Figure 3, applying industry-level shocks to all firms of a given sector amounts to underestimating the severity of the shock for some firms, and thus the overall share of insolvent firms.

Figure 3. Cumulated shares of insolvent firms over Mar-Dec 2020
(in % of initially solvent firms: impact of better accounting for firm heterogeneity)

“Fully simulated” means that shocks are sector-level shocks and take-ups are simulated.
Source: adapted from Hadjibeyli, Roulleau and Bauer (2021b).

The Committee on the Monitoring and Evaluation of Financial Support Measures for Companies Confronted with the Covid-19 Epidemic has also made use of firm-level data, e.g. showing that government support had benefited relatively more to smaller firms with intermediate profitability (see Coeuré, 2021). These firms are more likely to have avoided insolvency thanks to government support.

Stage 3 (since July 2021): financial statements and corporate defaults for 2020

These microsimulations based on observed data on payrolls and take-ups were still unsatisfactory. In particular, they were still using end-2018 financial statements. Due to the transition of a tax credit to a tax cut, French firms received an extra transfer worth 1.4% of their value added on aggregate in 2019. This double count came on top of robust GDP growth that year (+1.8%). Furthermore, monthly VAT returns are a noisy source of information. For instance, data for August and September are disturbed by accountant vacations; and very small firms only report quarterly or even yearly.

In May 2021, a subsample of 205,392 balance sheets for year 2020 (firms with annual turnover of at least 750 k€) were analyzed by the Banque de France, showing that only 14% of them had experienced both a rise in gross financial debt and a decrease in their cash position. In the fall of 2021, the French Treasury found a close proportion of 13% for a larger sample of 707,000 non-farm, non-financial firms subject to the “normal” tax regime and for which statements were also available in 2018 and 2019, representing 75% of turnover in France. It was found that, between 2019 and 2020, the median turnover variation was -5.7% while the median change in operating expenses amounted to -4.6 percent of pre-crisis turnover. Thanks to government support, though, the median financial result was reduced by only 0.3 percent of pre-crisis turnover between 2019 and 2020. However, compared to a year before, the distribution of profit variations was skewed to the left, meaning that a higher share of firms suffered from a drop in their profit rate in 2020 (Figure 4).

Figure 4. Annual variation in financial results
(in % of turnover)

Source: French Treasury (preliminary).

A complementary source of information on corporate health came from the number of business failures according to the records of commercial courts, which evidenced a sharp reduction in business failures in both 2020 and 2021, compared to 2019 (see Figure 5). Boekwa Bonkosi, Épaulard, and Gache (2021) have noticed that this decrease has been smaller for larger firms, leading to a more limited drop in the number of jobs in defaulting firms. They have also emphasised that the drop in defaults is roughly the same for those sectors most affected by the crisis as for other sectors, suggesting that public support has been instrumental in limiting defaults.

Figure 5. Monthly number of defaults relative to same month in 2018 and 2019 (average)

Reading: in March 2020, the number of defaults was 43% less than the average in March 2018 and March 2019. Defaults cover judicial settlements and liquidations. Dotted line represents the average over March 2020-October 2021.

Source: Maadini and Hadjibeyli (forthcoming) based on Bodacc.

At the end of 2020, Cros, Épaulard and Martin (2020) matched the data on defaults with pre-crisis financial statements in order to investigate a possible zombification of the economy. They found that the factors explaining defaults in 2020 were the same as usual, concluding to a “hibernation” rather than a zombification. Maadini and Hadjibeyli (forthcoming) further found that defaulting firms in 2020 displayed more deteriorated characteristics, notably lower productivity, than those defaulting in 2019, and that this deterioration was larger than for other firms, meaning that the cleansing effect was still at play, although with fewer defaults. However, their econometric exercise shows that the impact of financial characteristics was mitigated during the crisis, probably due to the interference of other sources of heterogeneity.

The prognosis of non-zombification made in 2021 was consistent with two other studies:

The Committee on the Monitoring and Evaluation of Financial Support Measures for Companies Confronted with the Covid-19 Epidemic (2021) found that “zombie firms” did not receive more support than their share of the economy;
Bach et al. (2021) found that state-guaranteed loans were not much extended to low-profit firms, and that net debt had not much increased for firms benefiting from these loans.

Finally, the impact of the crisis on the corporate sector was also apprehended through their bank accounts. Épaulard et al. (2021) have screened the (anonymized) bank accounts of around 100,000 small and very small firms up to August 2021. The proportion of firms with low net position was found stable in August 2021 relative to December 2019, although with varying situations depending on sector and location.

***

The Covid crisis offers an extraordinary experience for research in economics. The wealth of firm-level data will surely help to draw lessons that exceed this crisis and push the frontier of knowledge forward. After analyzing the phasing-out of government support, time will come for ex-post evaluation: has government support to corporations achieved its objectives? Was it cost-efficient? This last stage will be carried out by independent researchers. A lot of interesting work upcoming for the profession in the years to come!

***

>> Version française : Les données d’entreprises durant la crise Covid en France

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Poverty, purchasing power, employment: statistics tested in everyday life

2021-11-26T00:00:00+01:00

On November 3^rd, INSEE published a first estimate of the monetary poverty rate for 2020: 14.6%, the same level as in 2019, one of the lowest among European countries (see table 1 for the calculation of the poverty rate).

Table 1: Calculation of the income poverty rate

In autumn 2021, the median standard of living in 2020 is not yet known. In 2019 it was 1,837 €/month per consumption unit, according to INSEE. Assuming that it has evolved in 2020 in line with the average disposable income (+1%), it would be €1,855/month in 2020 and the poverty line would be €1,113/month per consumption unit. A quick reminder: this threshold is calculated after social transfers received (active solidarity income, in-work benefits, housing benefits, family benefits).

For sure, one cannot be satisfied with a poverty rate close to 15%. In France, the minimum wage (Smic) is relatively close to the median income, so that situations of poverty are often the result of unstable or part-time employment, combined with a difficult family situation and/or, sometimes, the non-use of benefits. According to the Conseil d'analyse économique (2017), means-tested social benefits make an important contribution to reducing income poverty. However, access to stable employment affects several dimensions of poverty, including social isolation (see the ATD Fourth World and Oxford University report, 2019 on the multiple facets of poverty, and INSEE surveys on material and social deprivation).

Poverty through two crises

In times of crisis, it is particularly important to protect the poorest households, who cannot draw on their savings to overcome temporary income losses. The exceptional support granted to low-income households during the pandemic contributed to this (see my December 2020 post). Indeed, the stability of the poverty and inequality rates during the 2020 crisis, if confirmed, contrasts with their increase during the previous crisis, whereas the fall in GDP was much more marked in 2020 than in 2009 (table 2).

Table 2. Two crises compared: 2009 and 2020

In his recent blog post, Jean-Luc Tavernier, the Director general of Insee, explores the reasons behind the gap between the statistics (preservation of purchasing power for all income deciles during the crisis, confirmed by the analysis of the anonymised accounts of more than 200,000 La Banque Postale customers), and the "perception" of households. In addition to the issue of undeclared income which, by construction, escapes public statistics, he suggests that for some people, poverty may have intensified without the numbers increasing.

One explanation could be that the crisis has frozen disparate situations. For example, the number of minimum income beneficiaries (Revenu de Solidarité Active - RSA) increased in 2020 due to an abnormally low number of exits from the system rather than an increase in entries (Figure 1). Staying longer on the RSA could have had an impact on the intensity of real poverty (exhaustion of resources) and perceived poverty (lack of prospects for recovery).

Figure 1. Monthly entries in and exits from RSA (in thousands)

Source : CNAF, Estimations avancées des évolutions des foyers allocataires du RSA, juillet 2021.

Statistics and perceptions

A similar gap between statistics and perceptions can be observed concerning the evolution of purchasing power over a long period. According to the OFCE, purchasing power per consumption unit will have increased by 1% per year on average over the period 2018-2022, i.e. the equivalent of €334/year per consumption unit (constant 2020 euros, see graph 2). For a family of four (2 parents, 2 children under 14), and cumulatively over 5 years, this represents an increase in purchasing power of €3,500. Why don’t people perceive this increase? An explanation may be found in rising house prices, which weigh on the budget of first-time buyers (20% of the population). Indeed, only rents are deducted from the purchasing power, but renters are only 40% of the population, and the corresponding index is less dynamic than apparent rents, which include an improvement in housing quality. Another factor is the sensitivity of low-income households to the recent rise in energy prices, which, according to the Family Budget survey, accounts for a larger share of the expenditure of the first quintile of the standard of living than that of the last quintile (the share of fuel expenditure being higher for the middle classes). Nevertheless, the impact of the rise in energy prices is greatly mitigated by various measures aimed at low-income households in particular: exceptional top-up of the energy voucher, partial freezing of gas and electricity tariffs, and an 'inflation allowance'.

Figure 2. Purchasing power of households per consumption unit

Source : OFCE, octobre 2021.

To understand the perception of households, we may also need to turn to behavioural economics. For example, Thaler (1980) and Kahneman and Tvsersky (1983) show that individuals behave differently depending on whether they are acquiring or disposing of a good, and that they also have a particular aversion to losses. If a public policy combines, as it is the case over the period 2018-2022, measures that are favourable to disposable income (gradual scrapping of the residence tax, increase in the in-work benefit, etc.), and unfavourable measures (increase in tobacco taxation, reform of housing allowances), it is possible that the unfavourable measures will be given greater weight in the minds of households than the favourable measures, even when the latter outweigh the former in quantitative terms for all deciles of the standard of living, and in particular for those at the lower end of the scale (Figure 3).

Figure 3. Redistributive balance of permanent measures implemented over the period 2018-2022

Employment

Poverty is not limited to the monetary dimension. Employment - particularly stable employment - which is strongly correlated with poverty in its various dimensions, is therefore probably a good indicator. The first estimate of employment in Q3 2021 confirms the strong rebound since the beginning of 2021. Employment has now exceeded its pre-crisis level, while the number of hires on contracts of more than one month (including permanent contracts but excluding temporary work) has also been rising for several quarters. In France, everyone knows people looking for a job, but also, increasingly, companies looking to hire. The labour market can be observed by everyone, and aggregate figures are known quickly. According to INSEE's monthly household survey, the fear of unemployment has fallen sharply since the beginning of 2021. The employment figures are therefore, if not consensual, at least understood in a relatively homogeneous way, which is the basis for a good debate.

Figure 4

Source : Insee, novembre 2021.

***

>> Version française : Pauvreté, pouvoir d'achat, emploi : la statistique à l'épreuve de la vie quotidienne

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Hot weather on prices

2021-11-12T00:00:00+01:00

Autumn is coming, temperatures are dropping. Before turning on the boiler, everyone will have a look at the spectacular rise in the price of gas (figure 1), which the Government has decided to freeze (for the regulated tariff, no pun intended) from October, considering that part of this rise will be transitory and that the price paid by the consumer can therefore be smoothed.

What remains are the past increases, which will affect households differently depending on their heating method. As a consolation, those who, like me, are heated with gas, will note that the price of wool is still about 30% lower than before the crisis. Rather than pushing the heat, we can buy woolens.

Figure 1. Natural gas prices in euros (TTF futures)

Source: Insee, Note de conjoncture, 6 October 2021.

Price signal or inflation

The year 2021 thus ends with a natural experiment in a double price signal: carbon-based energy (gas and, to a lesser extent, oil) increases in price relative to nuclear, hydro, solar and wind power; and the price of different energies increases relative to other goods and services. While some are trying to understand the interconnection of electricity markets or to calculate the impact of this double shock on carbon emissions, central banks are wondering whether the inflationary surge is sustainable or temporary. And there, surprise, there is a certain serenity in the euro zone, summarised by Isabelle Schnabel, member of the European Central Bank's Executive Board, in a recent speech: the 3.4% rise in the harmonised index of consumer prices in the euro zone in September 2021 compared with September 2020 (in France, +2.7% for the harmonised index and +2.2% for the index calculated by INSEE using a slightly different method) would only be a "sneeze" and not the start of a pneumonia. Indeed, the rise is not unprecedented in the young history of the euro area (figure 2). The ECB expects inflation to fall back below 2% as early as 2022.

Figure 2: Consumer price inflation (blue) and core inflation (yellow) in the euro area, year-on-year (%)

Source: European central bank.

The temporary factors that are currently pushing up inflation are well known: base effects (price rises are measured against exceptionally low levels in 2020), rapid and unexpected recovery in demand, disruption of freight, low stock levels, lack of wind in recent months, early winter in Russia (5°C in Moscow as early as September!), temporary production cuts (semiconductors, oil, gas, vegetable oils), chain effects in other markets (electricity). Thus, the price rises observed in recent months are not necessarily sustainable, even if the downturn should be faster for some supplies (gas) than for others (semi-conductors).

Prices and wages

In the 1970s, an increase in energy prices of 9.3%, as forecast by INSEE for 2021 in France (after a 6.1% drop in 2020), would undoubtedly have triggered an inflationary spiral: largely indexed to prices at the time, wages would have automatically risen, which would have increased production, and therefore consumer prices, and therefore wages, etc. And the French franc would have been devalued. Today, this chain of events is far from being mechanical, due to greater competition between producers (within each country and between countries), consumption that is still below its pre-crisis level and an irrevocably fixed exchange rate within the euro zone. The main question is the labour market: will the current pressure on recruitment lead to significant wage increases? For the time being, experts do not see a boom in this area in Europe, although they remain cautious.

The recruitment difficulties of companies are real, but so far they remain less than those that prevailed before the crisis, at least in France (graph 3). And at the time, this did not prevent above-potential growth in activity, for a moderate rise in wages. Moreover, according to the August 2021 Acemo-Covid survey (Dares), recruitment difficulties are only really a constraint in certain sectors such as education or health, with companies in other sectors citing input constraints or still demand shortfalls as the main factors limiting production.

Figure 3. Share of firms reporting recruitment difficulties (%)

Source: Insee.

Some inflation is better than no inflation

The general public is often surprised that economists define price stability as inflation of around 2% (or even 4%) and not 0%. While too much inflation is bad for the economy, moderate inflation is better than no inflation. It thus gives space to monetary policy, allowing the real interest rate to be clearly negative even if the nominal interest rate cannot be (at least not much). It also allows relative price adjustments to be made without plunging sectors or countries into deflation. This is particularly important in a monetary union because a country cannot recover competitiveness losses through devaluation. If it is kept under control, as expectations to date seem to indicate, the rise in inflation could thus encourage an adjustment of relative prices between sectors and between countries in the euro zone.

When he was President of the European Central Bank, Jean-Claude Trichet used to hold up a graph showing the comparative evolution of prices in the member countries of the euro area, sounding the alarm well before the 2009 crisis about the cumulative loss of competitiveness of certain countries in relation to others. During the crisis, some countries such as Ireland and Spain experienced a dramatic fall in their relative prices. After the crisis, however, the lack of inflation in the euro area reduced the room for manoeuvre for these adjustments. The current supply-side tension, which is observable in all countries, could be an opportunity to make relative price adjustments without too much pain within the single currency. From this point of view, the fact that recruitment tensions seem to be more marked in countries with a current account surplus (Germany, Netherlands, Belgium) than in countries with a current account deficit (France, Portugal, Slovakia, see graph 4) is encouraging.

Chart 4: Vacancy rates in Q2 2021, not seasonally adjusted, in %.

Source: Eurostat. In blue: Eurozone countries. * 2021Q1. ** Firms with at least 10 employees.

>> Version française : Coup de chaud sur les prix

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Where does the money come from?

2021-09-06T00:00:00+02:00

If you are an economist like me, perhaps you were bombarded by questions from your relatives and friends during the summer break, around the theme:

Where does the money come from?

While the government seemed to be running out of money, trying to reduce the deficit, carrying out reforms in this direction; it suddenly takes a U-turn, pouring billions of euros to finance masks, tests, vaccines, but also companies whose revenues slumped during the health crisis, and households whose incomes would have otherwise collapsed. Our companies had no more customers, and yet we were still getting paid at the end of the month. How is this miracle possible? What’s the catch?

Everyone knows the normal money flow in the economy: firms pay their employees for their work; salaries provide the bulk of their income to households, which spend most of it on goods and services produced by firms, which pay their employees, and so on. (Figure 1).

Figure 1: The normal money flow in the economy

During the first lockdown, in March 2020, consumer spending fell sharply, unevenly across sectors; downstream sectors (such as restaurants) reduced their purchases of intermediate goods (such as bottles of wine), and the fall in sales spread through the economy (see Barrot, Grassi and Sauvagnat, 2020). I will illustrate the point here with the case of France, but the patterns are qualitatively similar in the other advanced economies.

Figure 2: Household consumption expenditure in volume
(Index 100 as of Q4 2019)

Source: Insee, Quarterly Accounts (30/07/2021) and Economic outlook (01/07/2021). Forecasts from Q3 2021.

Deprived of sales, many companies could no longer pay their employees, further damaging the circuit shown in graph 1. To avoid this catastrophic amplification of the economic crisis, the government very quickly put in place three types of support:

Direct subsidies, notably through the solidarity fund and exemptions from social security contributions;
A large extension of the short-time work scheme;
State-guaranteed loans and deferral of contributions and tax payments.

This support has enabled companies to continue paying salaries. The purchasing power of gross disposable income thus increased by an average of 0.4 percent in 2020, according to INSEE, while GDP fell by 7.9 percent (Figure 3).

Figure 3: GDP in volume and purchasing power of gross disposable income (GDI)
Index 100 in 2019

Source: Insee, Economic outlook July 1, 2021. Forecasts for 2021.

Thus, the purchasing power of households' gross disposable income increased in 2020 and 2021, but their consumption remained depressed on average over the period. Household savings therefore increased sharply. These savings were mainly hoarded in liquid form: bank deposits and saving accounts (Figure 4).

Figure 4: Net flows of liquid investments by French households

Households could have lent this money directly to the government, but this was complicated and not very profitable, since interest rates were (and still are) negative. Instead, they mainly left it in their bank accounts or saving accounts.

How did the banks invest your money? Broadly speaking, they had four choices:

Keep the money as deposits in the banks' bank (the European Central Bank - ECB - via the Banque de France);
Buy French public debt securities;
Lend to French companies and households;
Lend to foreign governments and companies.

The first two investments carried negative interest rates, but that was the price to pay to keep the money safe. The banks thus spread their assets over the four types of investments.

However, the ECB sought to encourage banks to lend more to the private sector. To do so, it offered to buy back their government debt, not only short-term securities, but also long-term bonds. In doing so, it wanted to encourage them to lend to other counterparts than the government, while putting downward pressure on the interest rates paid by all borrowers - public and private. In fine, the objective was to prevent the Eurozone from falling into a deflationary spiral: to facilitate public intervention, to encourage households to consume and companies to maintain their investment programs; in short, to sustain a certain level of demand despite the restrictions and uncertainties.

Together commercial banks and the central bank therefore recycled household savings to lend to the government, which in turn helped businesses to continue to pay out income to households: with the normal circuit of spending blocked, the government, with the help of the ECB, set up a temporary detour (see Figure 5). By keeping interest rates very low, the ECB has also reduced the cost of additional government debt, while encouraging firms and households not to cut spending too much.

Figure 5: The money flow in the Covid economy

Over the whole of 2020, French households saved around €100bn more than in 2019. They also invested less, so that their financial savings grew by around €110bn. For their part, public administrations financed €73.7bn of exceptional measures (including €27.4bn for short-time work, €15.9bn for the solidarity fund and €14bn in exceptional healthcare expenditure - amounts measured in national accounts). The government also guaranteed €130bn in bank loans to companies and granted them numerous deferrals of contributions (see details in the report of the Committe on the Monitoring and Evaluation of Financial Support Measures, July 2021). On the revenue side, the government has suffered from the fall in activity, which has reduced the various tax bases almost proportionally (although compulsory contributions have shown some resilience to the crisis, falling less than activity).

In accounting terms, the general government deficit in 2020 (€212bn in the provisional general government account for 2020) and the residual financing needs of French companies (€30bn) were largely financed by households (€180bn), via the financial system, with the remainder (€60bn) met by foreign investors. The money created by the banking system (primarily the ECB) does not change this accounting. Indeed, the bank deposits of households and companies are comparable to a debt of the banks towards them; and when you withdraw 20 euros from a cash dispenser, you only transform one form of asset (a bank deposit) into another (a banknote), without modifying the total. Thus, the banks have "borrowed" from households and businesses, as much as they have lent to them. The ECB's massive purchases of government debt have inflated its balance sheet in a balanced way on the assets and liabilities side. In other words, the ECB has created as many means of payment as it has acquired government debt.

As the economy has reopened, households have returned to consumption and money has returned to its "normal" flow. However, everyone is coming out of the crisis with accumulated savings (households), accumulated debt (companies and especially the government), or both (commercial banks, ECB, see Figure 6). The key question will be how these stocks of assets and debts will influence the new flows of production and spending, and how they can gradually be deflated without stopping the recovery.

Figure 6: The money flow in the post-Covid economy

***

>> Version française : D’où vient l’argent ?

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

A good read for the summer: the National Recovery and Resilience Plan!

2021-07-16T00:00:00+02:00

Green light! On July 13th, the European Council approved France's National Recovery and Resilience Plan (NRRP), which will enable our country to receive €40bn of European funding as part of the Recovery and Resilience Facility, financed by the "NextGenerationEU" joint loan.

For the fanatics, I highly recommend the 815 pages of the French NRRP, which describe in detail how European funding will contribute to a higher, greener and more digitalised economic growth in France. Once you’re done reading the French NRRP, feel free to dive into the German, Italian, Spanish, etc. plans. At this stage, 24 Member States have submitted their NRRP: all of them except for the Netherlands, Bulgaria and Malta.

First reading

There are several ways to look at NRRPs. First, following, Darvas and Tagliapietra (2021), one can compare the volumes of investment, especially those related to the European priorities - greening the economy (minimum 37% of investments) and digital (minimum 20%). Useful from a macroeconomic point of view, this first approach is more complex than it seems, for several reasons.

First, some national stimulus packages - notably in Germany, France and Belgium - are financed in part by national resources, which come on top of European grants. Others, notably Italy, Greece and Romania, have requested European loans in addition to grants. The macroeconomic impact of the plans and their distribution by priority must then be assessed taking into account all those dimensions, including the national components (Figure 1).

Figure 1

Note: Amounts excluding Multiannual Financial Framework and REACT-EU funds. For the three Member States that did not submit their NRRP, the amount shown is the one they should receive according to the expected distribution of the Recovery and Resilience Facility.

Second, the different stimulus packages do not all cover the same period. In both Germany and France, stimulus measures are concentrated on the period 2021 to 2023 and are pre-financed at national level before being gradually reimbursed from European funds up to 2026. Conversely, Spain and Italy are expected to spread their investments out to 2026, mainly using the loan component of the RRF (see the case of Italy in Figure 2). Romania and Slovakia have even announced stimulus spending until 2030.

Figure 2

Source: Italian Stability Program, April 2021. Measures are accounted for using national accounts methodology which tracks the year expenditure took place; European instalment corresponding to these expenditures comes in with a delay.

Third, the distinction between emergency measures (during the crisis) and recovery measures (after the crisis) is not always clear. In both France and Germany, recovery programs were designed during the summer of 2020. They have been implemented while unexpected new health restrictions had to be imposed and have de facto supported firms and households during these difficult periods. It is therefore tempting to add the two types of measures together (Figure 3).

Figure 3

Sources: NRRP and national budget documents, French Treasury computations.

Finally, a comprehensive macroeconomic approach must also take into account the automatic stabilizers, which differ across Member States. For example, one can add up the additional deficits in 2020, 2021 and 2022 compared to 2019 (Figure 4), which mechanically feeds in higher public debt. According to this last metric, the cumulative effect of support and stimulus measures reaches nearly 20% of GDP for France, compared with 10% in the previous chart (discretionary measures only) and 4% in Figure 1 (discretionary measures of the recovery plan only).

Figure 4

Sources: Ameco, French Treasury computations.

However, the latter representation is still unsatisfactory. Not only is the cumulative variation in public balances endogenous to the severity of the crisis, but also and above all, NRRPs are intended to be long-term plans for transforming economies. The planned investments are not "simply" billions of euros injected into the economies, but they are also targeted public interventions aiming at raising the pace of potential growth and making the necessary transformation to face challenges associated to climate change. These investments come with a program of structural reforms that cannot be overlooked. Let’s start over.

Second reading

The French NRRP is presented as a succession of nine "components": energy renovation, ecology-biodiversity, green infrastructure-mobility, green energy-technologies, business financing, etc. For each component, the plan specifies "targets" and "milestones" - intermediate objectives that will trigger instalment from the European Commission according to a precise timetable. These targets and milestones may relate to investments or reforms. For example, the component entitled "Job preservation, youth, disability, vocational training" includes numerous investments in vocational training, support and hiring aids, but also reforms, notably that of unemployment insurance. Targets indicate quantitative objectives (e.g., 100,000 additional apprenticeship contracts by 2022), while milestones provide qualitative objectives (e.g., a first set of measures to reform the unemployment insurance scheme will come into effect in 2021, the remainder in 2022, provided economic activity is as good as hoped). The French NRRP includes more than 150 targets and milestones, 38 of which are to be achieved by 2021 (Figure 5).

Figure 5

Source: European Commission, Proposal for a Council Implementing Decision on the approval of the assessment of the Recovery and Resilience Plan for France. Should the targets for year N be assessed as completed the corresponding instalment is made in year N+1 (except for 2021 with a lump sum payment corresponding to 13% of the total amount of the plan). The 2022 instalment is expected after the 38 targets and milestones for the year 2021 are assessed as completed. The Commission has selected a total of 175 targets and milestones.

The reforms included in the NRRP are closely linked to the specific investments, to make them more effective. In some cases, they indirectly contribute to the effectiveness of the recovery plan by improving the medium-term fiscal outlook. Finally, as for every other Member states, France has to present reforms that also address the recommendations made by the European Union in the context of the European Semester. The number of reforms varies from one country to the next, but is generally higher in countries with the largest NGEU transfers as a percentage of GDP, Italy having a record of 508 of targets and milestones (Figure 6).

Figure 6

Source: NRRPs submitted, grants only.

At this point it is clear that the NRRPs are not Keynesian-style stimulus plans, but rather plans for the transformation and efficiency of European economies. There is much at stake for the European Union here. Indeed, we are talking about nothing less than the potential growth of the different countries (and therefore their capacity to bear high debts without risk of crisis), but also their capacity to face the challenges of ecological transition. Member states cannot disappoint their partners now that they have embarked on the path of common debt. Hence the tedious system of commitments and reporting set up. NRRP is also a good textbook to learn the concrete implications of issuing a common debt.

***

>> Version française : Lecture estivale : le Plan National de Relance et de Résilience !

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Labor Reallocations: Time for Practice

2021-07-08T00:00:00+02:00

According to the latest forecast from the Banque de France, the level of activity in France, in June is only 3% below its pre-crisis level, and this gap will be closed by the first quarter of 2022 (Chart 1). Does this mean the crisis is behind us? Not quite yet.

Looking for the omniscient worker

As we all know, GDP is the sum of firms’ value added. Some of them have already returned to their pre-crisis level of activity (some have even exceeded it), while others are lagging behind. Within sectors, the recovery is not uniform because demand has shifted. If workers were omniscient, versatile and perfectly mobile, they would instantly relocate to wherever there is a demand for labor. There would be no unemployment and no constraint on GDP if demand is there.

But of course the real world is a different story. A stewart, for instance, cannot turn himself overnight into a heat pumps installer in private homes or a care provider for the elderly. Not only does he usually lack the right skills, but he often ignores that such job opportunities exist. The "representative worker" producing units of GDP per capita does not exist. If demand picks up slowly in air transport and more strongly in thermal renovation, the former industry will be demand-constrained in the short run while the latter will be supply-constrained. As Restrepo (2015) has shown, difficulty in identifying the right workers out of many applicants may even lead firms in dynamic sectors to restrict hirings, which in turn reduces workers' incentives to train. In a nutshell, skills mismatch affects negatively the economy in a proportion that exceeds the size of the relevant labor force segment.

Source: Banque de France, Macroeconomic Projections, June 2021.

The idea of an uneven recovery is not theoretical. Compare the activity trajectories of sectors heavily affected by the crisis (Figure 2, top) to those which have been relatively unaffected (Figure 2, bottom). By April 2021, the sectors in the right-hand chart had returned to a level of activity close to (or even higher than) the pre-crisis level, while those on the left were still far from it. With the rebound from the reopening of the economy, the vast majority of workers in the sectors on the left-hand chart will return to employment. However, long-distance and business travel recovery will be gradual and perhaps incomplete, weakening the demand for air travel (and therefore airplanes), but also for taxis, international hotels, conference centres, luxury shops, etc. A major challenge in the coming period will be the ability of our economy to reallocate some of the workforce from these weakened sectors to expanding sectors, such as thermal renovation, care or digital-related activities.

Figure 2: Loss of activity compared to 2019 Q4, in %

Source: French Treasury.

Skills to be transferred

While the vast majority of job changes occurs within the same sector (see Davis and Haltiwanger, 1992), cross sector mobility is still a significant phenomenon, especially in accommodation and food services, where 9 percent of the workforce switches to other sectors on average over a 6-quarter rolling window (average over 2008-2018, see Figure 3). The fact that employees in this sector are often young, at the beginning of their professional career and hired on short term contracts may explain this relatively high mobility. In the manufacture of transport equipment, on the other hand, a sector that includes the automotive and aircraft industries, workers are much less mobile: 2.7% move to another sector during a similar 6- quarter window.

Figure 3 - Sectoral mobility in two sectors (over six consecutive quarters, average 2008-2018)

Manufacture of transport equipment

Accommodation and food services

Source: INSEE (continuous Employment survey, 2008-2018 editions), calculations: French Treasury.

Note: On average over the period 2008-2018, 2.7 percent of people working in transport equipment manufacturing had moved to another sector 6 quarters later. Among them, 1 percent moved to scientific and technical activities.

Each worker has a unique combination of skills, experiences, motivations and personal constraints. Some skills, such as proficiency in English, accounting, or digital technologies, are easily transferable across sectors; it is not so much the case for other skills such as sales or inventory management. Table 1, from Trésor-éco No. 287, compares skills most commonly used in job-destroying sectors in 2020 (on the left) to those commonly used used in the job-creating sectors over the same period (on the right). We can see, for example, that the ability to manage a team is required in 53% of jobs in "laying off " sectors and 72% of jobs in hiring sectors. Therefore statistically, more than half of the workers occupying at-risk jobs have a skill that is in high demand in growth sectors. In contrast, inventory management is a skill that is common in "laying off" sectors but not so much in hiring sectors.

Table 1. Skills most used in job-destroying sectors* and in job-creating sectors** in 2020

* Eight sectors that were the most job destructive in 2020: accommodation and food services, arts, entertainment and recreation, trade, administrative services, other services, transport equipment manufacturing, metallurgy, financial and insurance activities.

** Six sectors that were the most job-creating in 2020: IT activities, scientific research and development, medical and social accommodation, education, construction, human health activities.

Sources: Pôle Emploi ("node" skills), Insee (continuous Employment survey, 2019 edition).

Note: 53% of employment in the eight most job-destroying sectors uses the skill "team management". This skill is also found in the six most job-creating sectors, where 72% of employment requires this skill. The skills in green are those that are common to both job-destroying and job-creating sectors.

Using Italian data from 2015 to 2018, Basso, Grompone and Modena (2021) confirms that job proximity in terms of skills has a highly significant impact on the probability of job-to-job transition. To maximize the chances of workers in at-risk jobs to move to an expanding sector, a quick fix would consist in having them trained for the skills listed in blue in the table on the right. Unfortunately, this is where things get complicated.

First of all, the skills described in the table are quite broad, for example, knowledge of standards and safety varies greatly depending on the sector concerned. Therefore, the needs may actually differ from the specific skills of available workers.

Second, some skills are difficult to acquire, such as "generalist and specialist medicine" or, to a lesser extent, nursing and accounting. More generally, skilled workers are more likely to have the specific skills required in job creating sectors: on the left of the table 1, skills marked in green are often those of skilled workers, while those marked in blue are more often those of low-skilled workers. Unfortunately, job training opportunities are less often directed at low skilled workers, particularly in high-risk sectors such as retail.

Finally, it can be tricky to get workers in vulnerable sectors to train for new projects: aside from family constraints and the costs associated with possible geographic mobility, they also need to identify the interest in retraining, to be informed about the range of possibilities and how to access them.

Hence we cannot emphasize too much the importance of professional training in the coming period, particularly for low-skilled workers. There are already many schemes both for those in employment and the unemployed. The Recovery Plan has strengthened them to provide better support for employees benefiting from the furlough scheme or whose jobs are threatened as well as for young people and jobseekers to guide them towards promising occupations. The reform of professional training, in particular through the personal training account, allows for an easier access to training. It is essential to continue along this path, which implies above all effective information for the people who need retraining. In this area, innovative ideas and ever more dedicated action will be needed to ensure that schemes reach their target, keeping in mind the need for careful evaluations so as to continuously improve the schemes. This is just the beginning of a broader jobs reallocation that will be spurred by the ecological and digital transitions.

***

>> Version française : Réallocations de main d’œuvre : place aux travaux pratiques

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Supply and Demand, Season 2021

2021-06-03T00:00:00+02:00

Supply and Demand: 2020 in retrospect

At the beginning of the health crisis, an intense discussion immediately unfolded amongst economists on whether the economic crisis to come would be mostly the result of a demand shock or of a supply shock. The administrative closures of theaters, restaurants or sports halls were both a supply shock (the services no longer being supplied) and a demand shock (banning the access to this type of establishment). However, for sectors that were still open, the proportion of the two types of shock could vary: school closures, supply-chain disruptions and the implementation of health protocols in companies would constitute a supply shock - a drop in productivity; but the fall in demand for goods and services linked to closed sectors (wine, taxis, etc.) could also be significant, sometimes even drammatic, and could be further aggravated by the fall in household and business incomes, as modelled by Guerrieri et al. (2020) et Baqaee et Farhi (2020).

Thus, it quickly became clear that it was necessary to use a sector by sector analysis, also taking into account the interactions between sectors: for instance, a closed restaurant reduces demand for laundry services, the latter also being subject to supply constraints (health protocols, closed schools, etc.). For France, Dauvin and Sampognaro (2021) have relied on the INSEE employment survey, the DARES Acemo-Covid survey, input-output tables and a model describing the interactions between the 17 broad sectors of the economy. After aggregating the results, they found that supply-side constraints (administrative closures, absenteeism due to school closures, supply disruptions) were responsible for about 2/3 of the 31% drop in activity recorded in France in April 2020, the remainder being attributable to the drop in final demand.

Biden, or endogenous supply

A year after this traumatic experience, the debate over supply and demand seems to have almost disappeared, under American influence. More precisely, it has been transformed. On its own, the $1.9 trillion stimulus package in the United States (9% of GDP, which follows the previous administration's stimulus package and precedes the investment plan and the one for families and education) greatly exceeds the output gap of the US economy, estimated at around 3 percentage points by the Congressional Budget Office. The rationale behind this is the following: by making the US economy overheat, the limits of production capacity will be pushed up: part of the inactive population will re-enter the labour market, workers will train to develop skills needed to respond to a shifted demand, companies will invest more, etc. Of course, inflation could temporarily increase, but this is a minor problem and even an advantage in a period when it is so low. Conversely, accepting even a transitory deficit in activity could involve lasting damage to the economy through more or less persistent hysteresis effects (see Blanchard, 2018 ; Fatás and Saxena, 2020).

So why not do the same in France? Between 2019 and 2020, the French public deficit increased by around 7 GDP points ( if we correct for the double accounting of the CICE tax credit in 2019), while GDP decreased by about 8%. In the United States, the deficit increased by 9 points for a decline in activity of only 3.5%. Should France have done more? This is not obvious, given that the poorly targeted support measures in the United States have resulted in much greater increase in savings than in France (see Figure 1). The limited decline in employment in France (−1.5%, compared to −7.9% for GDP) does not seem to validate the need for more stimulus in the short run either. In 2009, employment fell by 2.2% while GDP fell by 3.8%. As we wrote in a previous blog post, public action in 2020 was devoted to preserving the production system (companies and workers) rather than to a Keynesian stimulus as such.

Figure 1: Cumulative increase in savings, difference to 2019 – GDP points

In 2021, the public balance should remain quasi stable both in France and in the United States, in percent of GDP. However, pre-crisis GDP will most likely be already recovered in the United States, unlike in France. Is the latter not stimulating enough? First of all, activity is still constrained by the pandemic in France, more than in the United States where health restrictions seem to have weighed less on the economy. Plus vaccination is also in the United States more than a month ahead of France. When parts of the activity are shut down, a boost in demand is unlikely to yield satisfactory results. However, vaccination is progressing and there is good hope that most of the constraints will be lifted by the summer. Why not follow the U.S. lead with additional and ambitious fiscal stimulus in the second half of the year? Could a strong rebound in demand drive an increase in potential GDP, so that there would be no significant supply constraints? The scenario is attractive, but two elements argue for some caution:

France entered the crisis with still a high level of unemployment (7.8% in 2019Q4 for metropolitan France, against 3.5% in the United States); however, the pressures on production capacities, and in particular hiring difficulties, were significant (see Dares, 2020). France is a much more open economy than the United States: imports of goods and services account for 33% of GDP in France, compared with only 15% in the United States. Thus, while excess demand in the United States can result in a surge in inflation and/or a return to work of the inactive, in France, the most natural way to serve additional demand is through imports. Moreover, a price increase would be more problematic in France than in the United States as France is part of a monetary union: the resulting losses in competitiveness will not be recouped through currency depreciation.
Also due to its membership to the euro area, France cannot table a debt trajectory as the one foreseen by the CBO for the United States (see Figure 2). Indeed, the euro does not (yet) equal to the dollar as an international currency: demand for euro-denominated assets is not infinite; even if it were, debt divergence with respect to other Member States could shift market perceptions, not to mention the risk of infringing European fiscal rules. Finally, the record of France is not as good as that of other countries when it comes to reducing its debt ratio in good times. In short, France has to be more cautious than the United States when it comes to public finances, even though, like other countries, it is currently benefitting from historically low interest rates.

Figure 2: US federal debt in % of GDP

Source: Congressional Budget Office, March 2021.

Season 2021

The gradual lifting of health constraints in May-June 2021 is going to unlock demand: consumers will return to the terraces of bars and restaurants, go shopping, travel, etc. Some supply constraints (such as absenteeism for quarantine) will also decrease. Consumption may even surprise on the upside, as it was the case in the summer of 2020. Does this mean that supply can easily meet any rebound in demand, no matter how high? Although this is a possibility, several difficulties must be overcome (see Fernald and Li, 2021).

The first obstacle will be the residual health restrictions, in particular the gauges that may still affect the productivity of establishments open to the public: if only half of the seats are available in restaurants and theaters, productivity will mechanically be almost halved.

The second obstacle will be to adapt the supply to changes in demand. Let us imagine, for example, that tourism recovers but becomes more rural and less international. Since hotels and their staff cannot instantly move from the large cities to the Atlantic coast, the Alps or the countryside in Normandy, supply could prove in excess in some places and insufficient in others. More generally, the persistent nature of some effects of the crisis, notably the attrition of long-distance tourism, but also the digital transformation of the economy (the rise of e-commerce and teleworking) will require inter- and intra-industry reallocations that will take time to materialise. The average shopping basket of a European teleworker or holidaymaker may also be smaller than that of a commuting executive or a Chinese tourist.

The third obstacle will be the capacity to use production factors - capital and labor - immediately at the same level of productivity as before the crisis, after long months of reduced activity. The loss of field experience and capital maintenance, as well as the greater difficulty in integrating young people into companies during the crisis, could affect productivity during the first few months of the recovery. There are also delays in recruiting seasonal workers, but also regular employees who may have taken advantage of the months spent in partial activity to prepare for another future (Les Echos, 10 April 2021).

A flexible script

Like most forecasters, the French Treasury adopted a cautious and non-definitive view on potential GDP, with the assumption of a slightly lowered level (-2 ¼ %, see Figure 3). Under this framework, GDP would reach its potential by 2024, and any additional demand stimulus would have less and less effect on activity as GDP (black line) moves closer to its potential (red line).

Figure 3: Potential and effective GDP

However, this estimate of potential GDP is highly uncertain. It cannot be ruled out that, after this enormous but perfectly exogenous crisis, the pre-crisis path will remain unchanged (pink line); just as a lower potential GDP cannot be ruled out. Under such circumstances, the fiscal strategy should be to adapt government support to the demand and the supply according to the evolution of bottlenecks: to stimulate the demand in the absence of pressure on the supply side; or the supply in the opposite case, with the emphasis on training workers and restarting businesses, and possibly a granular approach depending on the industry or on the territory.

In times of great uncertainty, the economic policy response must be first and foremost flexible, attentive to an economy whose fragility may come from supply or demand, investment or consumption, industries or territories, households or companies, aggregates or inequalities, etc. As the economy gradually reopens, it will be essential to scrutinize all available indicators with high frequency in order to calibrate the support as closely as possible to what is needed.

***

>> Version française : L'offre et la demande, saison 2021

>> Versión en español : Oferta y demanda : temporada 2021

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

3-6-12: French firms through the Covid storm

2021-04-20T00:00:00+02:00

Should you take home only three figures out of the just-released working paper by Benjamin Hadjibeyli, Guillaume Roulleau and Arthur Bauer on French firms during the covid crisis, these could be: 3-6-12:

3% is the proportion of French firms that would have become insolvent (meaning that the value of their debt would have exceeded the value of their assets) from March to December 2020 even without the crisis, while they were solvent before the crisis;
6% is the proportion of French firms that would have become insolvent during the same period, accounting for the impact of the crisis and the public support measures;
12% is the proportion of French firms that would have become insolvent during the same period, accounting for the impact of the crisis, if there had been no public support.

Hence, the crisis is expected to have roughly doubled the number of newly insolvent firms from March to December 2020. Without public support, though, this number would have quadrupled. Another way to understand these figures is that public support is likely to have halved the number of firms becoming insolvent over this period.

Of course these are rounded numbers. The exact simulation results are reported in Figure 1. Given the many assumptions that the authors had to make for this exercise, though, 3-6-12 can be accepted as an approximation.

Figure 1. Microsimulation results for French firms over March-December 2020

Reading: amongst all firms, 11.9% would have become insolvent over March-December 2020 without public support, a proportion that shrinks to 6.6% thanks to public support received, to be compared to the 3.6% in the counterfactual of no crisis. 71.1% of all firms remain solvent in all scenarios.

Source: adapted from Hadjibeyli, Roulleau and Bauer (2021).

Importantly, insolvency is a signal of financial distress but does not necessarily involve bankruptcy. As a matter of fact, 17% of firms (mostly young companies) were already insolvent before March 2020. As for newly insolvent firms, they may not be short of cash, or their debts may be restructured, or they may nevertheless receive additional funding when activity resumes, or they may be taken over by a group. Conversely, solvent firms can go bankrupt if they become illiquid (i.e., if they run out of cash to meet their current obligations). These distinctions will be crucial for the phasing out of emergency support, which will need to avoid both the risk of excess liquidations and that of a “zombification” of the economy due to the debt overhang.

Microsimulations

These new results are derived by simulating the financial statements of 1.8 million non-financial firms (82% of French firms value-added) over the course of 2020. Turnovers are measured through individual monthly VAT records. Firms react to negative shocks by reducing their payroll, through short-time work, non-renewal of short-term contracts or layoffs. Observed data are used up to June 2020 (turnover) or September 2020 (payroll). For the rest of the year, the simulations rely on sector-level data and on a number of assumptions. Specifically, firms are supposed to partially adjust their variable costs, whereas they cannot adjust their fixed costs.

Public support consists in short-time work (compensation for the wages of workers who cannot work – 84% of the worker's net wage with a floor at minimum wage, with firm-level take-up data up to September), direct subsidies through the “solidarity fund” (a subsidy scheme for SMEs particularly affected by the crisis, with firm-level data up to November), tax deferrals (only employer’s social contributions in the simulation, firm-level data up to October) and tax reliefs (employer’s social contributions for firms particularly affected by the crisis, fully simulated). When individual take-ups data are missing, the grants are simulated based on the eligibility criteria.

When costs exceed revenues, a firm is supposed to first exhaust its cash, and then to borrow. It is assumed that, thanks in particular to state-guaranteed loans, firms are not subject to credit constraints. Hence, they accumulate debts when they run out of cash. When debts exceed assets, a firm is deemed insolvent.

In the simulations, it is assumed that, without public support, firms would have had similar behaviour in terms of layoffs. In particular, they would have kept in their payroll the employees that in fact were put in partial activity. This assumption is probably extreme, since layoffs would have been much larger without the short-time work scheme. In the absence of any reliable guideline to estimate firms’ counterfactual behaviours, the choice was made to rely on an overly simple but transparent assumption. Hence, over March-December 2020, the 12% figure probably somewhat overestimates the number of French firms that would have become insolvent without public support: some of them could have survived by laying off employees. It nevertheless provides an order of magnitude of the proportion of firms at risk of massive layoffs, restructuring or (ultimately) liquidation.

Heterogeneity

Which form of public support has made the difference? According to Figure 2, large firms and mid-caps seem to have benefited mainly from short-time work: the proportion of newly insolvent firms is similar with the whole set of public support schemes (blue bars) as with short-time work only (hatched blue bars). In contrast, very small firms have benefited relatively more from the solidarity fund and from tax reliefs.

Figure 2. Share of newly insolvent firms by size

(in percent of all firms)

Source: Hadjibeyli, Roulleau and Bauer (2021).

Unsurprisingly, accommodation & food is the most severely hit sector (Figure 3). Without public support, almost 30% of firms in this sector would have become insolvent between March and December 2020. With public support, this share falls dramatically. Information & communication is in a very different situation, with relatively high insolvency rate in normal times and not much more as a result of the crisis.

Figure 3. Share of newly insolvent firms by sector

(in percent of all firms)

Source: Hadjibeyli, Roulleau and Bauer (2021).

Cleansing

The simulations also provide an insight on potential cleansing effects of firm exit. Figure 4 plots the distribution of labour productivity across initially solvent firms, after controlling for firms’ sector and size. In normal times, newly insolvent firms are on average less productive than the average firm: the red line peaks at a lower level of productivity than the black line. With the crisis, insolvent firms are still less productive than the whole sample, but the gap is reduced: the blue lines peaks in-between the red and the black ones. Hence, the cleansing effect is still there, but it is less effective than in normal times. Note that public support does not distort the distribution of newly insolvent firms: it does not systematically protect more productive nor less productive firms (solid and dotted blue lines are almost coincident), likely because the support was provided very broadly.

Firms have been affected differently by the crisis depending on their sector and size. For instance, firms in the accommodation & food sector are typically less productive than manufacturing companies, and smaller firms are generally less productive than larger firms. This composition effect could add to the cleansing effect and raise aggregate productivity. However, this is only a short-term effect. The distribution of activity across sectors after the crisis is still extremely uncertain. Furthermore, aggregate productivity will also depend on new entrants. Replacing food services by healthcare may not raise aggregate productivity.

Figure 4. Distribution of labour productivity (in log) across initially solvent firms

(controlling for sector and size)

Note: The sample is restricted to firms with at least one employee and excludes initially insolvent firms. The black line represents the distribution of labour productivity for the whole economy. The red line represents the same distribution for the subsample of firms becoming insolvent between March and December 2020 without a crisis. The solid blue line shows the distribution for the subsample of firms becoming insolvent with the crisis but without public policies. The dashed blue line shows the same distribution but accounting for public policies (short-time work, payroll tax deferral, tax relief, SMEs Solidarity Fund).

Source: Hadjibeyli, Roulleau and Bauer (2021).

Debt overhang

In the simulations, firms borrow either to roll over their maturing debt (in which case, the stock of debt stays unchanged) or to cover their new liquidity needs once they have used up their cash reserves (this borrowing increases the stock of debt). The additional debt may push a firm into insolvency, or just result in a deterioration of its balance sheet. Under this assumption, it is estimated that, from March to December 2020, non-financial firms would have borrowed €72 bn to cover their liquidity needs in normal times. This number doubles (€148 bn) with the crisis, accounting for the subsidies received.

Even though corporate investment is mostly driven by demand which is expected to rebound, a debt overhang can weigh on investment (see Kalemli-Özcan et al. (2019) for instance). Thanks to an econometric modelling of the reaction of corporate investment to firms’ indebtedness, the simulations suggest that, even if demand and profits get back to their pre-crisis level, the new debt caused by the crisis could translate into a 2% reduction in tangible investment compared to its pre-crisis trend. Various components of the French recovery plan aim at counteracting this effect to avoid long-term scars to the capital stock. In particular, permanent cuts on production taxes will support firms’ margins, while quasi-equity injections (prêts participatifs and obligations Relance) will contribute to strenghtening their balance sheets.

>> This work was carried out for and presented to the Committee for the Monitoring and Evaluation of Support Measures for Companies Confronted with the Covid-19 Epidemic chaired by Benoit Coeuré > Link to the report (in French)

>> Version française : 3-6-12 : Les entreprises françaises dans la tempête Covid

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Europe size 55

2021-03-23T00:00:00+01:00

In June, the European Commission will announce its 'fit-for-55' package, designed to put the European Union on a pathway to reduce net greenhouse gas emissions by 55% by 2030 (below 1990 levels), so as to achieve EU carbon neutrality by 2050. The menu is set to be substantial, with a starter, main course and dessert trolley.

For starters : the revision of the Emissions Trading System (EU ETS) for heavy industry, power generation and intra-EU air traffic, improving its functioning, possibly extending its scope (including to shipping and aviation) and introducing a Carbon Border Adjustment Mechanism.
The main course, then: an upward revision of the ambitions of sectors not covered by the quota market, which are responsible for around 60% of emissions: transport, construction, agriculture, waste, and some industries.
The dessert trolley, finally: revision of the Regulation on land use, land use change and forests (to capture carbon), revision of the Renewable Energy Directive, energy efficiency, energy taxation, revision of the Regulation on CO2 emission standards for light commercial vehicles...

In this menu worthy of the best restaurants currently closed, the EU's Carbon Border Adjustment Mechanism (CBAM) plays a useful, if not indispensable, additional role in strengthening the climate policy instruments. The idea is that producers from particularly emission-intensive sectors face de facto the same carbon pricing whether they are in the UE or not. Exporters into the EU, wherever they come from, would, for example, have to purchase emission allowances upon entering the EU. Companies that can demonstrate a virtuous production process or that are already subject to such a policy in their home country should be offered flexibilities so that these elements are taken into account when calculating the level of the adjustment, in order to encourage greater climate ambition.

The challenge of carbon leakage

According to the meta-analysis by Branger and Quirion (2014) and the literature review by Carbone and Rivers (2017), for every 10 tons of greenhouse gas emissions avoided in a country or region implementing more ambitious climate policies, emissions in the rest of the world would increase by 0.5 to 3 tons. These leakages can be either direct (relocation of carbon-intensive production, loss of market share by countries with ambitious climate policies at home and in the world) or indirect (via a drop in energy prices at the global level due to lower demand from countries with ambitious climate policies). If econometric studies have long struggled to identify carbon leakage, it is perhaps because ambitious policies examples were relatively scarce. Recently, it has been possible to document such leakage. In fact, according to the Commissariat Général au Développement Durable (2020), France's domestic emissions fell by around 20% over the period 1995-2019, but its carbon footprint increased over the same period by 7% due to a 72% rise in emissions associated with imports. This raises concerns about the effectiveness of European climate policies or, worse, the stalling of European ambitions themselves.

According to the European Court of Auditors (2020) the sectors at risk of carbon leakage (steel, cement, chemicals, paper, etc.) and benefiting from free allowances will account for 94% of industrial emissions during phase 4 of the EU Emissions Trading System (2021-2030). Calibrated on the most efficient installations in terms of emissions for each sector, these free allowances do not cancel out the incentives to decarbonise. However, they have been allocated in too large a number. It is difficult to eliminate the allocation of free allowances for industries that are highly subject to international competition and that suffer from unfair competition from countries with lower or no ambition in this area.

There are also difficulties regarding with public opinion. According to the survey conducted in France by Thomas Douenne and Adrien Fabre (2019), French people are on the whole very much opposed to the idea of using carbon pricing in order to change behavior and thus fight climate change. They are skeptical about the effectiveness of such a mechanism, which they see essentially as a pretext for raising taxes, or even as a license given to the wealthiest households, who will be able to buy their way out of any constraints. The idea of a Pigouvian tax the revenues of which would be redistributed to low-income households or even to the middle class - an idea supported by many economists and in particular by the Conseil d'analyse économique and the Conseil des prélèvements obligatoires - is also rejected by the respondents. They individually seem to underestimate the net gain that such a policy could bring them, even when they receive precise information about it. Only the tax on kerosene seems to get their support, perhaps because most of them believe they are not subject to it, at least not directly. Conversely, they are in favour of norms and regulations, even though their implicit cost is often higher than the one of a tax. They also support public investments, the ultimate financing of which remains unclear and which are generally considered by economists as a complement rather than a substitute for pricing (see Bureau, Henriet and Schubert, 2019). In general, the real and perceived fairness of climate change policies seems to be a sine qua non condition for its success. In this respect, international burden sharing is a key element, as Europe accounts for less than 10% of global emissions. Indeed, carbon pricing in Europe already appears high in international comparison (see figure below), although it will need to increase further if the EU is to meet the Paris Agreement targets.

Graph: Status of carbon pricing in the world in 2020

Note: The levels of carbon pricing and territorial emissions coverage are those of 1^st November 2020. For the EU, its ETS and the carbon taxes implemented by its Member States, including France, are taken into account through a weighting. Local and regional initiatives are taken into account for China, Canada, the United States and Japan.

Source: L'Heudé, Chailloux and Jardi (2021), French Treasury based on data from World Bank (2020). Carbon Pricing Dashboard.

Don't lose sight of the real target

If the introduction of a CBAM is necessary in order to strengthen European policies against global warming (see L'Heudé, Chailloux and Jardi, 2021), those who dream of making foreign companies bear the costs of our policies will be quite disappointed.

To be compatible with the rules of the World Trade Organisation, the CBAM must be non- discriminatory, i.e. it must apply the same treatment to imported goods as to European products. Thus, carbon pricing could be based on that of the EU ETS and take into account the climate policies of third countries and their level of development. It will also be necessary, inevitably, to phase out the free allocation of allowances to European industry.

Furthermore, the CBAM will automatically increase the costs of downstream companies: if imported steel has to pay for carbon allowances, then the European automotive industry, which consumes steel, could suffer from competitiveness loss compared to its non-European competitors on third markets and even on the internal market (as long as this downstream sector is not itself covered by the CBAM). In addition to this effect, European exporting companies will suffer the loss caused by the gradual phase out of free allowances. In a recent study, Cecilia Bellora and Lionel Fontagné (2021) estimate that a CBAM affecting imports of the most carbon-intensive goods (in particular steel, cement, chemicals, aluminum, refining, paper and glass) at the level of the European pricing would reduce European exports by around 1.5%, which raises the question of compensation.

This is why the implementation of a CBAM can only be gradual and accompanied by intense diplomatic action to ensure that as many countries as possible (particularly the United States, which recently returned to the Paris Agreement) implement ambitious climate policies, which will make it possible to adjust as closely as possible the scope of the CBAM and, if necessary, to deal with its undesirable effects.

Ultimately, the implementation of a well-calibrated CBAM should incentivize third-country producers to engage in low-carbon production. The CBAM is neither a means to make others pay, nor is it designed to improve the competitiveness of European companies. Its role is to strengthen the effectiveness of climate policies and to enable better international coordination on the subject. And its success will ultimately mean… its demise.

>> Version française : Europe taille 55

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Whatever it costs, how much is it?

2021-02-26T00:00:00+01:00

Trip to America

The announcement of a rescue plan to support the US economy of USD 1.9 trillion (i.e. about 9% of GDP, on top of 4% decided in December) has raised controversy over the proper size of emergency and stimulus plans. For Larry Summers and Olivier Blanchard, yet fervent supporters of expansionary fiscal policies, this programme is too massive because it is between 2 and 4 times the output gap expected for 2021: with a unit multiplier, its implementation would boost demand far above the country's production capacities. Inflation would start to rise again, causing the Federal Reserve to hike rates, lessening the impact of the fiscal stimulation, in line with the good old IS-LM model, or worse, triggering a financial crisis. The inflationary risk would even be higher since the savings accumulated by American households since the start of the crisis is approximately twice the amount of the stimulus plan envisaged by the Biden administration: if half of this savings were consumed, that would double the stimulus provided by this new relief package.

This view is vigorously contested by several economists including Gauti Eggertsson and Paul Krugman. For them, the goal is not to bring the GDP as close as possible to its potential level, but rather to speed up vaccination, reopen schools, provide support to households weakened by the crisis as well as to states and municipalities to dissuade them from laying off their staff and shutting down public services (see Figure 1). All the funds will not necessarily be spent (it will depend on the evolution of unemployment) and even so, it is not that obvious it would trigger inflation as the estimates of the output gap and its link to inflation are uncertain. And after all, wouldn't it be good news if inflation started to rise again?

Figure 1. Components of the Biden rescue plan

Source : Krugman (2021)

Paul Krugman does not speak of a stimulus package, but rather of a ‘disaster relief plan’, or even of war spending, which cannot be decided on the basis of a macroeconomic analysis. Although inflation has often accompanied war efforts, he stresses that beneficiary households and communities will not hesitate to spare some of the funds received, thereby reducing the pressure on prices. Blot, Dauvin and Sampognaro have calculated that public transfers have overcompensated the income losses of American households in 2020 (see figure 2), without noticeable effect on inflation.

Figure 2. Contribution to household disposable income at the end of the third quarter in 2020

(First three quarters of 2020 relative to same period in 2019)

Source : Blot, Dauvin et Sampognaro (2021)

For Blanchard and Summers, aid to victims could be better targeted and at least partially funded by the winners of the crisis. They do not worry about the size of the Biden plan, but about its concentration on current spending in 2021, when the country critically needs to catch up in public investment, over several years. They warn against the illusion of inflation being gone for good. A few years ago, they recall, negative interest rates were inconceivable. You can never tell.

For Blanchard and Summers, the relief plan must be adequately calibrated, but not too generous, so as not to cause inflation or windfall effects. If it turns out to be insufficient, it will always be possible to increase it. Eggertsson and Krugman want to strike people's minds with a massive plan to suddenly pull the country out of the rut by reversing expectations and accepting windfall and inflationary risks.

Note that none of them raises the issue of the additional debt that will arise from the plan. They probably consider that a further increase in debt by 9 points of GDP is painless as long as the interest rate is lower than the growth rate. They therefore assume a zero opportunity cost for public money in the country of the dollar.

Back home

Is the Biden relief plan good news for the Eurozone and for France? The IS-LM model in an open economy (or Mundell-Fleming model) gives us a straightforward answer: “it depends”. The stimulus plan will indeed trigger U.S. imports that will benefit our businesses; but a rise in long-term interest rates in the United States could contaminate other parts of the world, notably the euro area. Ultimately, it will go well if the dollar appreciates, as suggested by the standard model, which does not take into account, for example, the risk of the Fed being subject to fiscal dominance (see Jean-Pierre Landau).

Should the euro zone or at least France follow suit? By setting an upper limit not to the public debt ratio, but to the “likely” interest payments relative to GDP, Xavier Ragot estimates that there is room for a second French stimulus plan of € 100 billion (around 5% of GDP). The reasoning is neither Krugman's “whatever it takes” nor Blanchard and Summers’s “whatever is needed”. It is rather a matter of "as much as possible": take advantage of low rates and use all available space to borrow and finance capital expenditure, which will be spread over several years.

What does that say? For 2021, French GDP is still expected to remain below its 2019 level, with the gap ranging between -3% and -4% according to forecasters. Before the crisis, the European Commission forecast for France a growth of 1.3% in 2020 and of 1.2% in 2021. Without Covid, the GDP of 2021 should therefore have been 2.5% above its level of 2019. The total loss compared to this counterfactual would therefore be in the range of 6 to 7% of the GDP, i.e. around € 140 to € 160 billion. One might therefore think that an additional stimulus plan of € 100 billion would not only be welcome, but also be insufficient. This reasoning is however inadequate, for three reasons.

The first reason is the weakening of the level of potential GDP: the counterfactual of 2.5% above the 2019 level is inaccessible in the short term because the fall in investment and productivity have weakened the supply of goods and services at least temporarily. The right short-term objective is therefore to return to the production level of 2019. The second reason is that the government's recovery plan (‘France relance’) - € 100 billion mainly for 2021 and 2022 - is ramping up and it is far too early to consider it to be insufficient. The third and main reason is that the health situation is not yet normalized: when restaurants, theaters and other sports halls are closed, it is vain to believe that the activity can regain the level of 2019 by increasing demand. The most severely hit sectors (transport, accommodation and catering, arts and entertainment) account for 10% of GDP. Considering these sectors, which were operating at ¾ of their capacity in the 4th quarter of 2020 and ⅔ in November, will keep operating at low levels during the first half of 2021 and would rise in the second half of the year to a level close to that of the third quarter of 2020, the loss of GDP for these sectors alone compared to 2019 would be in the range of 2 ½ points, i.e. about half of the overall GDP loss, currently estimated at -5% by the Banque de France, or -4% in January by Insee.

This gap of some 2 ½% cannot be closed until the health crisis is behind us. However, for most forecasters, the French economy will likely return to its 2019 level next year. If, once the health constraints are lifted, recovery proved slower than expected – which can be closely monitored thanks to high-frequency data that are now swiftly made available - , then an additional boost to demand would appear adequate. For now, the 2020 experience rather shows that economic rebound can be quite sharp once health constraints are alleviated. Sound risk management would advocate in favour of keeping the additional fiscal space to face new unforeseen events. The low level of interest rates does not invalidate the reasoning, not to be mistaken with the long-term public investment strategy whose objective is long-term growth rather than short-term recovery.

In the immediate future, the best indicator to assess the effectiveness of public action on the economic front is the situation of businesses and households in their diversity: for several more months, the "whatever it costs" must be considered from a microeconomic perspective (accelerate vaccination, protect businesses and households weakened by the crisis) rather than from a macroeconomic perspective (restore the level of GDP). Things will change when health constraints are finally lifted.

>> Version française : Quoiqu’il en coûte, c’est combien ?

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

2021, zombie year?

2021-01-25T00:00:00+01:00

According to the Banque de France, the gross leverage of non-financial corporations increased by € 174.5 billion in France, including € 120.7 billion in State-guaranteed loans (PGE) between February and September 2020. This additional leverage, which represents just over 7 percent of 2019 GDP, adds to the pre-crisis debt-to-GDP ratio of 73.5 percent (see HCSF, 2020). The PGE is a flexible loan - firms can reschedule it for up to 5 years and benefit from low interest rates. It is nonetheless a debt that firms will eventually have to pay off. They may then have to sacrifice investment in order to secure repayment: the economic literature (see for example Kalemli-Özcan, Laeven and Moreno, 2020) emphasizes that a debt overhang weighs on investment, and therefore on growth. Some firms will not survive: according to Cros, Epaulard and Martin (2020), a company's debt ratio is a major determinant, alongside its productivity, in predicting future failure.

Schumpeter against the PGE?

From a sharp Schumpeterian perspective of “creative destruction”, the Covid-19 crisis could regenerate the French productive fabric by eliminating the least productive firms, freeing capital and labor for more innovative and productive businesses. Contrary to popular belief, the process of creative destruction works rather well in normal times in France (Ben Hassine, Le Grand and Mathieu, 2019; David, Faquet and Rachiq, 2020).

However, the Covid-19 crisis hits businesses in an undifferentiated way. As early as June 2020, OFCE warned of the risk that an abnormal proportion of highly productive businesses could fall into insolvency as a result of the crisis. In general, economists have validated the choice of many European countries, including France, to support businesses based on administrative closures and loss of turnover, with no condition on their pre-crisis financial situation. Inevitably, non-viable businesses were protected during the year 2020. Have they been over-protected?

Zombification not proven so far

The OECD qualifies as zombie a company aged ten years or more whose annual results fall short of interest payments, so that the debt increases steadily. These firms cannot invest in productive capital, let alone R&D; they survive and become a drag for growth. Will the legacy debt left by the crisis inflate the number of zombie firms in France, when, according to the OECD, their number was relatively limited before the crisis, by international comparison? Will public support - solidarity fund, short time work in particular - just mask the non-viability of a large number of firms for a time, without solving their root problem?

The simulations carried out by the French Treasury (see box below) do not validate such a scenario. Taking into account both the decline in activity by sector during the year 2020, and the main public support mechanisms for more than 700,000 businesses employing workers and viable before the outbreak of the crisis, we observe that the debt accumulated due to the crisis shifts slightly to the right the distribution of businesses at risk of insolvency: firms at risk of insolvency due to the crisis (blue line on the graph) are on average slightly more productive than those that would have been at risk in normal times (red line). However, those firms becoming insolvent due to the crisis still display lower productivity than average (black curve): firms with lower productivity are indeed the most at risk. Finally, public support (solidarity fund, short-time work), while reducing the number of firms at risk, do not modify their distribution in terms of productivity: the blue solid line (with support) is very close to the blue dotted line (without support).

Simulation results: distribution of labor productivity

Interpretation: each line represents the distribution of productivity within a sample of firms.

Source: French Treasury.

These simulations confirm the diagnosis of Bach et al. (2020) for the Institute of Public Policies, according to which the Covid-19 crisis has hit relatively unproductive businesses harder, with a marked sectoral effect (hotels and restaurants, transport, culture, etc.). It has also affected, in fewer numbers, more productive businesses that should not have faced an insolvency problem under normal conditions. If less productive firms have received more support on average, it is because they have been more severely hit by the crisis rather than due to a bias in public support in their favor: the aid has made it possible to drastically reduce the number of businesses at risk of insolvency, without selecting the most productive nor the least productive.

According to Cros, Epaulard and Martin for the Economic Analysis Council, corporate failures fell sharply in 2020 compared to 2019, but within failing businesses, the same factors (mainly productivity and debt) explain the failures in 2020 as in 2019. This confirms the neutrality of the support mechanisms deployed in 2020: the “zombification” of the economy is not there yet.

The year of danger

The indebtedness of French businesses at the end of 2020 is difficult to interpret for four reasons. The first is that, while leverage was high and growing in percent of GDP before the crisis, it was not necessarily the case in percent of total assets, especially for SMEs: the increase in debt was indeed accompanied by a parallel increase in corporate equity (see HCSF, 2020). The second reason is that businesses have accumulated cash almost equivalent to their additional debt since the start of the crisis; at the aggregate level, their net indebtedness has therefore only very slightly increased. The third reason is that businesses are also indebted to the state and social security. As of mid-November 2020, nearly € 23 billion in social contributions remained owed by French firms, according to the Business Aid Monitoring Committee. The last reason, finally, is the great heterogeneity of businesses, even within the same sector: businesses that have accumulated cash are not necessarily the same as those that have accumulated debt.

How to avoid a zombification of French businesses? Existing work on the topic identifies two key factors: restructuring procedures and the health of banks.

Efficient restructuring procedures: Jordà et al. (2020) like the OECD (2017) consider the speed and efficiency of restructuring procedures as a key element in avoiding "zombification". In particular, early procedures are strongly encouraged. In the French case, Epaulard and Zapha (2020) show that, for two similar businesses, the safeguard procedure (before cessation of payments) produces much better results than judicial reorganization. The current transposition of the European directive on restructuring and insolvency provides an opportunity to strengthen the system by shortening the safeguard procedure, developing early warning systems and confidential procedures (conciliation, ad hoc mandate), providing reliable statistical information on survival rates and, finally, by facilitating the adoption of debt restructuring plans with an increased role of creditors compared to shareholders and incentive voting rules.

Healthy banks: since the study of the Japanese crisis in the 1990s (Caballero, Hoshi and Kashyap, 2008), economists have considered that well-capitalized banks are less likely than others to maintain credit lines to non-viable businesses because they can provision non-performing loans without risking appearing under-capitalized. From this point of view, we must welcome the recapitalization effort of French banks since the 2008 crisis: the CET1 capital ratio of the main French banks rose from 6% on average in 2008 to 14% in 2019 (HCSF, 2020). The deployment of the PGE nonetheless raises the question of the selectivity of loans: distributed widely, it could artificially keep low-productivity businesses alive, which would slow down the recovery after the crisis by blocking the reallocation of production factors towards more productive businesses. For Schivardi, Sette and Tabellini (2020), though, the idea that supporting fragile businesses would come at the expense of dynamic businesses does not hold water. Indeed, the crisis entails two simultaneous effects: increasing the share of fragile enterprises and reducing the growth of healthy ones, without the former being necessarily the cause of the latter. Moreover, it is still too early to judge which businesses are promising. Finally, we must not overestimate the capacity of the economy to reallocate production factors in the midst of an economic crisis and even after the crisis, as Patrick Artus (2020) reminds us.

What conclusion should we draw from all this? First, an effort is desirable to promote preventive, confidential or public restructuring procedures, still too rarely used in France. Second, banks capitalization must be protected, in particular by limiting the distribution of bonuses and dividends, as the European Central Bank once again recommended on December 15, 2020. Finally, for another few months, the Schumpeterian argument will have to wait: because uncertainty will remain as to the lasting impact of the crisis on business models, the objective will still be to limit liquidations in order to preserve the productive fabric and skills. Businesses will have to seize the opportunities offered to reschedule PGEs, open their capital to new investors (in particular via the « Label Relance » funds) and repair their balance sheets with future quasi-equity loans.

French Treasury simulation method (December 2020)

The simulations presented above are based on the latest known balance sheets (Fare 2018) of 728,706 French businesses employing paid staff, to which sectoral shocks are applied based on INSEE data (17 sectors including trade, accommodation-catering, transport): quarterly national accounts for the first three quarters of 2020 and, for the last quarter, the estimates of the economic outlook of Insee of November 17, 2020, which therefore include the deterioration of the economy during the second lockdown.

The operating costs identified as variable costs (supplies) are supposed to adapt to the shock with a certain time lag, the calibration of which changes during the year 2020. Meanwhile the costs identified as fixed costs (in particular the payroll, taxes and rents) remain constant.

The short-time work scheme is supposed to cover the payroll in a proportion equivalent to the loss of activity (no retention of labor). For the solidarity fund, eligible businesses are identified at individual level and the amount awarded is determined from the amount actually paid by sector. These schemes act as subsidies and raise profit.

The financial statement of each company is updated by simulation every quarter. A company whose cash flow becomes negative is supposed to go into debt to cover its liquidity gap, which raises its leverage. A company whose debt exceeds the amount of its assets is considered insolvent. The concept of insolvency is different from that of default: a company may be insolvent but not be in default.

Without an economic shock, some businesses would nevertheless have become insolvent due to an unsustainable financial structure: the approach adopted here consists in comparing, at the end of 2020, the characteristics - and in particular the distribution of productivity - of firms becoming insolvent due to the crisis to those firms that would have become insolvent even without a crisis (insolvent firms at the start of the simulation being excluded from the analysis).

The methodology will be detailed in a forthcoming Treasury working paper by Benjamin Hadjibeyli, Guillaume Roulleau and Arthur Bauer. This technical document will also present a new set of simulations in which the economic shocks will be assessed at the level of each company and the effective take-up of the various support mechanisms (solidarity fund, short-time work, deferral of social security contributions) will also be taken into account at individual level.

>> Version française : 2021, l'année des zombis ?

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Public support to low-income households in covid times in France

2020-12-28T00:00:00+01:00

In many ways, the pandemic has affected households more severely at the lower end of the income scale: they frequently work in sectors particularly affected by the crisis, such as restaurants and transport; they often work on short-term contracts, and their occupations can hardly be teleworked; finally, they had to face additional food expenses when school canteens were closed. While no one disputes this reality, quantification remains difficult.

Struggling with the statistical apparatus

Insee's economic outlooks have now accustomed us to tracking consumption, mobility and electricity consumption on a weekly basis. After the first lockdown, it quickly became clear that consumption was rebounding strongly. During the second lockdown, it was quickly known that the economy would be less penalized. The success of the Black Friday was also disclosed quickly. Such acceleration of macroeconomic analysis comes in sharp contrast with the delays of the statistical apparatus in social matters.

The Department of Research, Studies, Evaluation and Statistics (DREES) of the Ministry of Health and Social Affairs has just published an estimate of the change in the number of beneficiaries of the social inclusion benefit (RSA) since the start of the crisis. From now on, it will publish a monthly update, which will speed up the identification of social distress. According to DREES, between October 2019 and October 2020, the number of RSA beneficiaries has increased by around 8.5%. This figure is in line with the decline in activity in 2020, estimated at -9% by the latest forecasts of the Banque de France in Dec. 2020. In a November note, the family branch of the Social-security system, however, attributed the increase in the number of beneficiaries mainly to a sharp slowdown in exits from the system: for obvious reasons, existing beneficiaries have had more difficulty than in the past finding a job. While being a long-term beneficiary of the RSA is certainly not an enviable prospect, the figures so far do not seem to show that a larger number of households have fallen into precariousness as a result of the crisis than is usually the case.

Another way to look at the situation of low-income households is to rely on surveys. The first wave of the EpiCov survey, carried out jointly by Insee, Inserm and DREES (135,000 people surveyed in May 2020), indicates that, among the poorest 10% of households, 35% declare a deterioration of their financial situation during the first lockdown, against 23% for all those questioned and only 16% for the top decile. The survey also highlights great inequality by age: young workers were hardest hit by the first lockdown. However, these ratios do not take into account the extent of the loss of income or, by construction, the emergency assistance received between May and November (see table below). According to the latest Insee Conjoncture in France, aggregate gross disposable income of households fell by 3.3% in the first half of 2020 but could rebound by 4.1% in the second half. Thus, according to Insee, over the year as a whole, household income would be preserved on average.

Table. Emergency assistance to low-income households in France, year 2020

To overcome the lack of high-frequency social data, the Council of Economic Analysis (CAE), in cooperation with a large banking network, has studied anonymized consumption data paid by credit cards as well as the evolution of 300,000 bank accounts between February and August 2020. Assuming that the first decile of the population was also, before the crisis, the first decile in terms of spending on credit cards, the CAE shows that the consumption of low-income households fell less than the average during the first lockdown and that it rebounded more afterwards. Consistently, the savings of low-income households (measured by the change in their different bank accounts net of changes in debt with the same banking network) decreased, while wealthy households saved.

Absent of high-frequency data on household income, however, it remains difficult to quantify in "real time" the impact of the crisis on the situation of low-income households. Indicators such as the drop in the number of over-indebtedness applications at the Banque de France on the one hand, and the lengthening of the lines at food banks on the other hand, make it difficult to draw up a coherent picture. The decline in informal work may have contributed to blurring the statistical picture, but such assumption is by definition difficult to validate and quantify.

**An ex ante analysis by micro-simulations**

In the absence of "real" data observed during the crisis, we must fall back on ex ante analysis by micro-simulations. This is what the Institute of Public Policy (IPP) has done, using the TAXIPP 2.0 model. First, the IPP has shown that low-income households are the most exposed to the crisis because of their professional situation, their occupation and their sector of activity. Then it has simulated the impact of the emergency assistance measures on households income, based on 2019 income. The IPP finds that the various targeted transfers (see table above) have raised the average annual income of the first twentieth in terms of living standards by more than 5%. They have raised the income of the next twentieth by 2.6%, the effect diminishing rapidly for higher twentieths. The IPP also finds marked redistributive effect of the solidarity fund (which has supported individual entrepreneurs), while short-time work has rather supported the middle class. All in all, the support measures appear to be heavily targeted to households with the lowest living standards.

These results are consistent with the micro-simulations carried out by the French Treasury based on the Saphir 2019 model, even though the two exercises are not identical. According to these simulations, emergency assistance mentioned in the table has raised the average living standard of the first decile by 2.4%, and by 1% that of the second decile. It should be reminded here that the standard of living adjusts the disposable income for the composition of the household. The figure below shows that the support was well targeted towards the most precarious households. Over a third of the households benefiting from the support were in the first decile of the population in terms of living standards, and nearly two-thirds in the first two deciles.

Figure - Impact of the emergency support on households’ living standards, by deciles of living standard before the crisis
(in % of living standards in 2019)

Source: Saphir 2019 model, enriched with 2020 legislation, excluding variations in income and measures related to Covid-19.

Reading note: the emergency measures (excluding students) have increased the living standards of households in the first decile (both beneficiaries and non-beneficiaries) by an average 2.4%.

Excluded from the analysis, students have been particularly affected by the crisis: many of them have lost their part-time job or the internship. Targeting support to them is particularly difficult because socio-fiscal data does not capture intra-family transfers in favor of students – although these transfers constitute more than half of their income on average. The emergency assistance that they received in June and November was targeted towards students benefiting from social aids and scholarship holders.

Other measures have been taken in the context of the crisis to fight poverty, such as strengthening emergency accommodation or support to NGOs. Now, the French recovery plan foresees a wide range of tools to foster integration of young people into the labor market (hiring bonuses, subsidized jobs, extension of civic service). Employment remains the best bulwark against poverty, and a delayed entry into employment are likely to weigh on new generations of workers for years.

>> Version française : Ménages modestes : impact des mesures de soutien exceptionnelles

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Lockdown #1– Lockdown #2

2020-11-09T00:00:00+01:00

When faced with a new problem, one tried and true method is to return to a known problem. How much will the November lockdown cost in terms of GDP losses? Let's look at what happened in April and think about it on the margins.

April’s standard

The first lockdown lasted 8 weeks between mid-March and mid-May. In the middle of this period, the month of April constitutes a reference point since the whole month was concerned, with rules invariant during the month. It was a very strict shutdown. The economic consequences were not long in coming: activity plunged by 30% compared to the pre-crisis period, before a fast recovery in end-May, June and July (Figure 1).

Figure 1 - Monthly GDP scenario

Source : Insee, calculs DG Trésor.

Unsurprisingly, the fall in activity was very uneven across sectors in April: very strong in restaurants, accommodation, air transport, the automotive industry, culture and recreational activities, construction; much less in agribusiness or information and communication (Figure 2). These different sectors have varying weight in the economy and the result was a 30% decline in activity compared to the pre-crisis period.

Figure 2: activity losses by sector in April 2020, in % compared to normal

Source : Insee

The Dares Acemo-Covid survey identifies 4 major causes of the decline in activity in April: administrative closures, loss of business opportunities, unavailability of staff and, finally, supply difficulties. These difficulties have played out in different ways depending on the sector. For instance, restaurants and retail trade suffered mainly from administrative closures and travel restrictions, while their suppliers (such as wine, textiles or business services) lost outlets. The construction sector suffered particularly from supply difficulties (construction materials and protective devices). Finally, non-market services were constrained by the absence of staff forced to stay home for health reasons.

Projections for the second lockdown

The second lockdown announced on October 28th differs from that of April in four major aspects. First, schools will remain open, which should allow parents of young children to continue their activity on site or by remote working. Second, remote working is now a well-established practice in companies, and employees are equipped: according to Dingel and Neinman (2020) around 37% of jobs are “teleworkable” in France. Third, sanitary protocols are also in place at production sites, and protective equipment is available: activity should be preserved in sectors such as construction or manufacturing. Finally, public services will remain open and business travel will not be restricted.

These new conditions should mitigate the impact of this new lockdown compared to that of April. To assess their intensity sector by sector, the French Treasury has analysed the results of the Acemo-Covid survey and applied an attenuation coefficient for each constraint. For instance, the impact of administrative closures is assumed to be similar to what was observed in April for the accommodation-catering, retail trade and recreational activities. Conversely, personnel and supply constraints are greatly alleviated in all sectors, while the losses of outlets are still present but somewhat attenuated. These new constraints are then applied to the activity losses observed in April to obtain the expected losses for each sector in November compared to a normal situation. The result is summarized in figure 3.

Figure 3: Impacts of lockdown#2 (y-axis) compared to lockdown#1 (x-axis)

Source: French Treasury, using data from Insee and Dares. The size of the dots represents the weight of the different sectors in GDP.

The sectors on the oblique line are those for which comparable losses are expected in November compared to April. These may be sectors hit hardly by the restrictions (accommodation and catering) or, on the contrary, little affected sectors (agriculture). The areas above the line are those that should suffer less in November than in April. We note in particular the case of construction, whose activity had fallen by 60% during the first lockdown, and which could decline "only" by 30% during the second lockdown. No sector is below the oblique line, which means that no sector should be more affected than in April, although heterogeneity can be strong within a sector.

The last step is to multiply each loss of activity by the corresponding sector's share of GDP. The contributions from the sectors are then added to obtain the aggregate impact of the second lockdown for a month of restrictions: around −20% compared to a normal month, against −30% in April.

2020 at −11%

The INSEE figures for the first three quarters indicate achievement carry-over of "growth" of −8.3% for the year 2020: if the 4^th quarter were stable compared to the 3^rd, the growth for the year would be -8.3%. Unfortunately, the second lockdown should lead to a decline in GDP in the 4^th quarter compared to the 3^rd.

In October, the economy was operating about 5% below normal. In November, the 20% drop compared to normal (due to second lockdown) therefore amounts to a 15% drop compared to October. A month of lockdown followed by a period during which severe restrictions would continue to weigh on the economy would result in a reduction of about 2.5 points in annual growth. The growth forecast for the year 2020 is finally obtained by adding the carry-over of the 3^rd quarter (−8.3%) and the expected loss in the 4^th quarter (−2.5%, October having been stable compared to September ), so around −11% in total for the whole of 2020 compared to 2019.

>> Version française : Confinement-Reconfinement

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

How to stimulate consumption

2020-10-21T00:00:00+02:00

Travelling the last few miles

Published on October 6, the latest Insee Conjoncture in France confirms the rapid rebound in household consumption over the summer, the consumption level in Q3 being "only" 3% below pre-crisis level. Since September, though, new health restrictions have interrupted the recovery in sectors that had not yet caught up (hotels and restaurants, transport, leisure), while the catching-up effects are fading for the consumption of durable goods.

Figure 1. Household consumption forecast (Insee)

(100 in 2019 Q4)

Source : Insee, Octobre 6 2020.

This recent development is consistent with the Government's forecast which also takes into account a decline in consumption at the end of 2020 (see post from September 22nd, Growth at the end of the tunnel). The question therefore is how to find the best way to "cover the last few miles" to bring consumption back to its pre-crisis level, at least in those sectors not subject to administrative constraints.

Consumption support can be achieved through two channels:

Income support, particularly for low-income households, whose propensity to consume is higher;
An incentive to consume rather than save.

Income support

Figure 2, taken from the 2021 Economic, Social and Financial Report, details how households’ income was kept almost constant during the “great lockdown”. In particular, employees have benefited massively from the short-time work while individual entrepreneurs have been able to call on the solidarity fund.

Figure 2. Contributions to households’ disposable income growth in the second quarter 2020 (Year-on-year)

Sources : Insee (final results, T2 2020), DG Trésor

Over the whole of 2020, we expect households’ disposable income to remain stable on average, with a marked deterioration for households in self-employed or precarious jobs, which were the first affected by the jobs’ losses.

The recovery plan includes several mechanisms to support household income. First, the exceptional short-time work scheme will be extended for sectors subject to restrictions, while long-term partial activity agreements, which are currently taking off, will benefit to sectors hit hardly such as aeronautics. Secondly, the solidarity fund will continue to be accessible and will be extended to the very small businesses most affected by the crisis, their social and tax charges will also be reduced. Third, the back-to-school allowance, paid in August, has been increased by 100 euros to support low-income families, and scholarship students benefit from the 1 euro meal voucher.

Should the government do more? According to the latest data from the Banque de France (L’impact de la crise du Covid-19 sur la situation financière des ménages et des entreprises – août 2020), between March and August 2020, households’ cash holdings and bank savings increased by € 111 billion in total, while their bank borrowings increased by only € 25 billion. The additional net bank savings are therefore € 87 billion. Including life insurance, which suffered an outflow of € 7.8 billion over the period, net savings since March are close to € 80 billion.

According to an estimate made this summer by the OFCE, all income deciles experienced forced savings during lockdown, and the disposable income of the first two deciles remained stable overall. However, we can expect a sharp increase in inequality within these first two deciles. Supporting households hard hit by the crisis is an objective in itself, even if the transfers paid are not entirely consumed (some households, for example, may prefer to repay their debts).

Incentivizing consumption

The second way to stimulate consumption is to offer a temporary benefit such as a temporary VAT cut (Germany), subsidized restaurant meals (United Kingdom), or "eco-vouchers" (see, e.g., the proposal of MEDEF in June 2020). The idea is to temporarily encourage consumption, even if it means accepting a setback when the scheme expires. In the French recovery plan, the expansion of the already existing system Ma Prime Rénov’ follows this logic, in the area of investment (energy renovation).

The temporary VAT cut in Germany could boost consumption, provided retailers pass it on to their prices, which they seem to have done little so far (but it's a bit early to conclude). Across the Channel, the Eat Out to Help Out system allowed the British to dine in restaurants on weekdays by paying only half the bill for a month. Though the direct effect was an immediate 27% increase in consumption from Monday to Thursday, the measure also resulted in a drop in the number of weekend customers (-21%) according to the Center for Economic and Business Research.

In France, the experience of reduced rates (especially in restaurants) is not encouraging. Rather than being passed on to consumer prices, these VAT cuts have mostly benefited corporate markups (see Benzati and Carloni, 2019). While higher markups can help stimulate investment and employment, this goal is being pursued in France by lowering production taxes, which is a much more direct measure.

In Germany as in the United Kingdom, the measures implemented to stimulate consumption are costly because they are not targeted. Another option could be to distribute "eco-vouchers" or "green vouchers" with limited validity to low-income households, allowing them to buy goods and services from a restricted list of environmentally friendly consumption. Such a measure would simultaneously target three objectives: supporting the purchasing power of the poorest, reviving consumption and greening the economy.

In 2009, Belgium set up such an eco-voucher scheme distributed by companies to their employees using a model similar to holiday or restaurant vouchers. The green check takes the form of € 10 vouchers, capped at € 250 per year, which can only be spent on a list of goods labeled as durable. The logic is similar to some already existing French support schemes, such as support for thermal renovation or the purchase of electric bicycles.

The advantage of the green voucher over French aid is that it covers a wide range of products, which mitigates the risk of price increases. The downside is the complexity of implementation if you do not want to introduce distortions between products and between distribution channels. First, the list of eligible products must be defined, in cooperation with professionals and NGOs, and in accordance with European Union rules of level playing field (for example, it is impossible to favour local consumption or purely national labels). Then, retailers need help to adapt their labelling, checkouts and payment systems, all for a limited period since the measure is set to be temporary. The risk is a delay and significant implementation costs, and therefore an insufficient and late recovery.

Furthermore, a temporary scheme will have negligible impact on the environment. The different objectives come into conflict here. The purchasing power of poor households is better supported by means-tested transfers than by the distribution of green vouchers, especially if the latter are, like restaurant vouchers, intended for employees.

Beyond low-income households, relaunching consumption primarily involves reducing uncertainties: controlling the health situation, cushioning the negative effects of the crisis on employment. The main challenge is therefore the rapid deployment of the employment policies provided for in the recovery plan. Greening consumption is another objective that must addressed by specific policies and followed over the long term.

>> Version française : Comment stimuler la consommation

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Growth at the end of the tunnel

2020-09-24T00:00:00+02:00

Beyond the fog

Economists are routinely debunked for their forecasts which generally prove to be wrong, especially in times of turbulence. It would be wise not to make any forecast. As a matter of fact, and contrary to popular belief, most economists are not in the business of forecasting. This was the case with me until recently.

However, the very purpose of macroeconomic policies is to stabilize the economy: they need to tighten when the economy is booming, and to loosen in times of crisis as it is the case today. To at least try to be counter-cyclical, monetary and fiscal policies need to rely on assumptions about the economic outlook.

Furthermore, the diagnosis about the state of public finances depends fundamentally on this outlook. Tax revenues are more or less proportional to GDP, so the tax-to-GDP ratio is not much dependent on the economy: in the event of a crisis, tax revenues fall but they stay more or less stable in percent of GDP. In contrast, public spending hardly reacts spontaneously to changes in GDP. In France, public spending represents around half of GDP. If the latter decreases by 10% in 2020, as the Government anticipates today, the deficit will automatically increase by 5% of GDP. Such huge deficit is not worrying as long as it is a one-off, and given that financial markets are willing to lend the corresponding amounts. However, members of Parliament must be properly informed.

Compared to the projection made in June for the Budget Orientation Debate, the government's new forecast for 2020 and 2021 takes into account three novel elements:

The relatively strong rebound in activity during the summer in France, and therefore a better than expected first semester;
The resurgence of the epidemic in France and in other European countries, which could curb consumption and output over the second semester, before gradual normalization in 2021;
The recovery plan announced on September 3, of which some measures (hiring bonus for example) are already in place, and which could raise the level of GDP by around 1 ½ percentage point in 2021 compared to a scenario without a recovery plan.

The combination of these three elements leads to forecast a 10% decline in GDP in 2020, followed by an 8% rebound 2021. In 2021, GDP would lie 2.7% below its 2019 level. Beyond their forecasts for the 2020 dip and the 2021 rebound, the most recent forecasts for France all place 2021 GDP between 2 and 4 percent below its 2019 level (Fig. 1).

Figure 1. GDP path, September 2020 forecasts
(index 100 in 2019)

Sources: Banque de France, Consensus Forecasts, OECD & French Treasury, September 2020.

Any economic projection exercise is based on a macroeconomic model (Opale, in the case of the French Treasury) in which a series of assumptions are injected on (1) the international environment (growth of the other countries, world trade, oil price, exchange rate, etc.), (2) the extra-economic national context (demography, health situation, climate change, etc.) and (3) public policies (Fig. 2). Usually, the uncertainties mainly come from the international context and from the behavior of households and firms. Today, though, all projections are heavily dependent on the very uncertain evolution of the epidemic, on how private agents (firms and households) and public ones (municipalities, universities, etc.) will adjust to this unprecedented situation and, finally, on how they will seize the various supportive schemes of the recovery plan.

FIgure 2. The making of forecasts

Recovery engines

The economic rebound in 2021 will rely on three drivers: household consumption, corporate investment and public investment.

Household consumption

In 2020, household incomes were almost entirely preserved thanks to several support mechanisms, in particular short-time work. Given the constraints weighing on consumption during the lockdown, households as a whole have accumulated savings that could reach about €100 billion by the end of the year. In fact, the household savings rate (ratio between savings and disposable income) jumped during lockdown, from 15% in Q4 2019 to 27% in Q2 2020 (INSEE). Over the year, the savings rate could stand at 21%, before dropping to 17-18% in 2021. The savings rate would therefore remain above its pre-crisis level due to precautionary behaviors related to the still uncertain health situation and to the deterioration of the labor market relative to 2019. For the same reason, household investment (particularly in housing) should remain below its 2019 level.

Figure 3. Household savings rate
(in % of disposable income)

Sources: Insee, French Treasury (September 2020).

Corporate investment

In 2020, firms were severely hit by the health crisis. Over the first two quarters, their markups (the ratio between gross operating surplus and value added) fell by more than 7 percentage points for non-financial corporations (Insee). Over the year as a whole, the markups for non-financial firms would decline by 4 pp relative to 2019. This decrease, which includes 1.3 pp due to the impact of the scrapping of the CICE (replaced since 2019 by social security cuts), appears relatively contained in view of the economic shock , and is achieved thanks to the numerous public support measures, notably short-time work. Not surprisingly, investment fell sharply in the first semester. Over the year as a whole, the fall in the corporate investment rate (the ratio of gross investment to value added) would be limited to only 1 percentage point as a result of the simultaneous, dramatic drop in both the numerator (investment) and the denominator (value added).

In 2021, a sharp rebound of markups can be anticipated, notably as a result of the recovery plan (in particular the cut in production taxes). The recovery of markups, combined with dynamic demand and equity injections (allowing fragile companies to repair their balance sheets) will support corporate investment.

Figure 4. Markup rate of non-financial corporations
(in % of added value)

Sources : Insee, French Treasury (September 2020).

Public investment

Public investment decreased during the first semester of 2020. This evolution is not only due to the crisis, but also to the local electoral cycle, since public investment is largely decided at municipal level and is traditionally low in election years. Over the year as a whole, however, public investment should decline much less than corporate and household investment, already playing a stabilizing role in the sinking economy. The recovery plan would contribute to a recovery of public investment by more than 9% in 2021.

This macroeconomic scenario, on which the 2021 budget will be based, will be detailed in the Economic, Social and Financial Report which will be made public in early October.

Read more :

>> Version française : La croissance au bout du tunnel

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

Boosting the economy high and far

2020-09-10T00:00:00+02:00

Why a recovery plan?

An astronaut returning from a space mission would no doubt be incredulous at how severely the world economy has deteriorated in just a few months. The crisis is particularly dramatic in developing economies. Within advanced economies, France is one of the hardest hit countries due to a severe lock-down in the spring and its strong specialization in highly affected sectors (aeronautics, tourism).

On August 28, INSEE published its new growth estimate for the 2nd quarter of 2020: –13.8%, after –5.9% in the first quarter. The fall is sharp - much worse than in 2009 (Figure 1).

Figure 1. Real GDP

(at previous year chained prices, seasonally adjusted and corrected for working days)

Source : Insee, Informations Rapides, August 28 2020.

However, the crisis is of a completely different nature to the Great Financial Crisis, and the policy responses are also different.

The 2009 crisis was clearly a debt overhang crisis: excess leverage of poor American households, of American and European banks, and of some European governments. The crisis was mitigated by expansionary macroeconomic policies and by an international commitment not to erect protectionist barriers. However, deleveraging is a slow process. Furthermore, the Europeans withdrew their macroeconomic support too quickly, which produced a second dip in 2011-12.

After the 2009 crisis, significant efforts were made to re-regulate the financial system and recapitalize the banks. The latter are now better equipped to deal with possible defaults from the borrowers. In addition, governments and central banks have deployed huge means to support businesses and households during lockdowns (see here for the case of France). These efforts have paid off: the rebound in consumption was spectacular in France during the summer (Figure 2), and corporate bankruptcies have so far been contained.

Figure 2. Monthly household consumption of goods

Volumes at the previous year prices (chained since 2014)

Source : Insee, August 28 2020

However, the recovery is uneven across sectors and it has slowed down in late summer. With contaminations rising again in the fall, the economic outlook for the end of the year remains highly uncertain. Unfortunately, other risks also loom over the world economy, e.g. on international trade.

What recovery plan?

The objective of the French recovery plan is to bring the economy back as quickly as possible to a growth trajectory that should be as close as possible to its pre-crisis path, represented by the dotted line in Figure 3. The short term and long term are linked insofar as the short-term crisis produces injuries that may prove to be lasting: companies in difficulty as a result of the crisis are not necessarily the least productive,[1] the lack of investment in the short term will constrain production capacities in the medium term, unemployed workers may experience deskilling, etc.[2]

After plummeting in 2020, GDP should bounce back in 2021. Based on past crises, though, the spontaneous rebound will unlikely bring activity back to the dotted line (see for example the 2009 crisis on the figure 1). Moreover, the recovery will be slowed down by persistent uncertainties on the pandemic front and on the labour market. Certain sectors remain constrained or even closed; the household saving rate, which jumped during lock-down, could remain at a higher level than before the crisis due to precautionary behaviors; and companies will lack visibility to spontaneously embark on ambitious investment programs.

Figure 3. Macroeconomic stimulation : short-term and long-term

GDP before and after the Covid-19 crisis

Source : French Treasury

It is in this context that the recovery plan was unveiled on September 3. The idea is to tackle the economic crisis at its root - in the accounts of companies - so that they can retain their workforce, hire young workers, invest and modernize. Before the crisis, the National Productivity Council pointed out the weaknesses of the French production system: skills mismatch, digital lags, insufficient return on research and development, heavy production taxes.[3] These weaknesses did not disappear with the crisis. We must therefore take advantage of the recovery plan to correct them, while helping companies transform their growth model to make it compatible with the energy transition, without sacrificing their competitiveness. The objective is both to stimulate the rebound in 2021 and to raise the level of GDP and - why not - its growth rate in the long term.

***

Read more :

>> Version française : Relancer l'économie haut et loin

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

[1] See Mattia Guerini, Lionel Nesta, Xavier Ragot and Stefano Schiavo, "Firm liquidity and solvency under the Covid-19 lockdown in France", OFCE Policy Brief n ° 76, July 2020.

[2] See Valerie Cerra, Antonio Fatás and Sweta Saxena (2020), “Hysteresis and the business cycle”, CEPR Discussion Paper 14531, March.

[3] National Productivity Council, First Report, July 2019.

Equity gaps in the French corporate sector after the great lock-down

2020-08-25T00:00:00+02:00

The lock-down

The Covid-19 pandemic represents an unprecedented shock to operating accounts. Depending on the sector, firms find it difficult to maintain their normal levels of production due to a combination of (1) administrative closures, (2) decline in productivity and (3) fall in demand. The drop in turnover, combined with the existence of fixed costs, translates into lower cash-flows and possibly into losses and defaults.

Faced with this situation, the French government has implemented two types of measures:

Liquidity support: deferrals of taxes and social security contributions, State-guaranteed loans;
Solvency support: short-time work, solidarity fund.

In the short run, both types of support improve firms’ liquidity; in the longer term, though, only the latter improve their solvency.

Figure 1 shows that, during the eight weeks of general lock-down (March to May), public support could have absorbed around 95% of the shock suffered by French firms: for a total of € 95 billion in lost value-added, the the drop in cash was only € 5 billion.

Figure 1

Sources : French Treasury

Of course, business situations are highly heterogenous. France has around 3.8 million companies according to INSEE, including around 250 large companies and 5,600 mid-size companies (ETI).[1] The overwhelming majority of businesses in France are SMEs, the vast majority of which are very small companies (less than 10 employees). These companies can be very vulnerable to a lock-down shock. By simulating the evolution of the balance sheets of almost all French companies (INSEE-FARE database), the Treasury has found that short-time work and the solidarity fund could have reduced by around 75% the number of companies turned illiquid (i.e. unable to meet their deadlines) by the lock-down, and this even before accounting for State-guaranteed loans. The methodology is detailed in the box below.

The economic crisis

However, the shock for companies is more long-lived than the lock-down itself, and moreover the tax and social security liabilities will need to be repaid. From March to June 2020, French firms have borrowed heavily, in particular by resorting to State-guaranteed loans (more than 500,000 companies have benefited from these loans) but also, for larger firms, by borrowing on the bond market. According to the Banque de France, corporate gross debts (excluding tax and social security liabilities) increased by € 152 billion during this period. On average, companies kept most of the borrowed funds (i.e. € 142 billion) in liquid form, probably in anticipation of future difficulties.

Over year 2020, the Treasury estimates that the public sector could absorb around 2/3 of the aggregate cost of the shock, leaving households and businesses to share the remaining. This is where the risk of insolvency comes in: having increased their leverage during the lock-down, a number of companies will be at risk of insolvency (debts greater than the assets). Assuming a 6-month shock (strict lock-down followed by a gradual resumption of activity over the rest of the year), and without additional support, business failures could increase by around 70% compared to the usual number of failures (52,000 in 2019 according to Altares).

The equity gap

Beyond insolvency, the challenge of the coming months will be that of investment: without investment, the French economy could enter an anemic spiral in which low investment would gradually reduce production capacity, which would weigh on employment and income and therefore on consumption, which in turn would discourage investment, etc.

Economic research[2] has shown that more leveraged firms tend to invest less especially during a crisis. Highly leveraged but viable firms will therefore need recapitalization.

For SMEs and midcaps, the Treasury estimates equity gap at around € 10 billion for a few tens of thousands firms. Of course, these estimates will have to be adjusted depending on the path of the recovery and on the provisions of the recovery plan (in particular with regard to cuts in production taxes).

The € 10bn amount may be compared to households’ financial savings of nearly € 4,000 bn, including new savings accumulated during the lock-down which may amount to € 100 bn by the end of the year. Households will likely remain reluctant to take risks. However, it is the job of financial intermediaries to offer savings products combining limited risk-taking and a higher return than those on sovereign bonds - currently negative up to the 10-year maturity.

Insofar as the crisis could kick viable firms out of the market, public intervention is justified. Indeed, these companies would not necessarily be replaced by more productive firms, as it is the case in principle in normal times ("creative destruction").[3]

Box: methodology

The microsimulation model of the French Treasury aims at examining the consequences of Covid-19 on illiquidity (when a company runs out of cash) as well as insolvency (when the asset / debt ratio falls below 1).[4] The model is based on FARE 2017 data from Insee, which lists the balance sheets and income statements of almost all companies in France (n = 3.8 million companies). After excluding a number of

companies for which the information cannot be used, the model restricts the analysis to nearly 2 million companies representing 91% of the value-added of the French economy.

The aim is to measure the impact, on firms’ balance sheets, of sectoral shocks as estimated by INSEE during the Covid crisis.[5] Public policies put in place by the government are also simulated: deferrals of taxes and social security contributions, short-time work, solidarity fund (however excluding State guaranteed loans). The model accounts for fixed costs (such as rents), while other costs are supposed to partially adjust to the shock. Specifically, the wage bill and input purchases are assumed to adjust in proportion to the shock suffered by the company.

Illiquidity occurs when the initial cash position is depleted. Beyond this point, the firm is assumed to borrow, which increases the risk of insolvency and therefore the risk of default.

***

>> All posts by Agnès Bénassy-Quéré, chief economist - French Treasury

[1] Mid-size companies have 250 to 4,999 employees and a turnover not exceeding € 1.5 billion (or a balance sheet not exceeding € 2 billion).

[2] See e.g. Kalemli-Özcan, S., Laeven, L. and D. Moreno (2018), « Debt overhang, rollover risk, and corporate investment: evidence from the European crisis, » NBER Working Paper, No. 24555, November.

[3] See Blanchard, O., Philippon, Th. and J. Pisani-Ferry (2020), “A new policy toolkit is needed as countries exit Covid-19 lockdowns”, Peterson Institute for International Economics, Policy Brief 20-8.

[4] Several studies have also relied on micro-simulations: Gourinchas P.-O., Kalemli-Özcan, S., Penciakova, V. and N. Sander (2020): “COVID-19 and Business Failures”, UC Berkeley Working Paper, forthcoming; Guerini, M., Nesta, L., Ragot, X., and S. Schiavo (2020), “Firm liquidity and solvency under the Covid-19 lockdown in France”, OFCE Policy Brief No. 76; Demmou, L., Franco, G., Calligaris, S., and D. Dlugosch (2020), “Corporate sector vulnerabilities during the Covid -19 outbreak: assessment and policy responses,” OECD Tackling Coronavirus Series. Despite different methodologies and coverage, the results are relatively convergent.

[5] See “Point de conjoncture of 23 April 2020” or “Point de conjoncture of 27 May 2020”, Insee.