Employment and the jobless rate are both expected to rise in 2021, but the size of these movements is very uncertain. The rise in employment is likely to be limited, while the upturn in the jobless rate risks being big. The France Relance recovery plan will surely help boost employment. Uncertainty over the size of its rebound is linked in part to the vigour of the economic recovery. Above all, employment recovery will be hampered by several headwinds: the lagged impact of the GDP plunge in 2020, the increase in corporate bankruptcies, persistent sector differences, the return to work of furloughed or short-time workers, and corporate efforts to restore productivity gains and margins. As to the unemployment rate, the dynamics of employment and the labour force are both uncertain. There is also the question of the profile of the increase in the jobless rate in 2021. Will it be a continuous increase or a bell-shaped curve? The most likely scenario is the first one, with a sharper increase in the first half that eases in the second half. The French labour market is unlikely to return to good health in 2021 (as defined by the government in its unemployment insurance reform), but the year 2022 seems like a more realistic horizon.
UNEMPLOYMENT RATES IN FRANCE AND THE UNITED STATESAfter a stand-out performance in 20191, the French labour market’s performance was even more remarkable in 2020 given the massive recessionary shock that swept the French economy. In part 1 of this article, we conducted a review of this very peculiar year in which the Covid-19 crisis had a much smaller impact on the French labour market than on GDP growth2.
What should we expect in 2021? Will the labour market continue to surprise on the upside? Although we cannot rule out this possibility, for now, prospects seem to be more mixed.
Payroll employment is expected to rebound, but the unemployment rate is also expected to pick up. And the size of these moves is highly uncertain. To be more precise, the balance of risks leans towards a mild increase in employment and a big rise in the jobless rate. This article will review the support factors and headwinds currently at work.
PAYROLL EMPLOYMENT IS SET TO REBOUND, BUT BY HOW MUCH?
Concerning the outlook for employment, there can be little doubt but that the situation is poised for a rebound in 2021. The big question is by how much? We can reasonably expect the year-onyear increase to be positive by the end of 2021. But will employment rebound sufficiently over the course of the year to return to pre-crisis levels by year-end 2021? Will the annual average growth rate swing back into positive territory? This would already be a significant improvement. Will it be strong enough to close the gap with job destructions in 2020 (when payroll job destructions averaged 328k)? Although we can answer the first two questions in the affirmative, the third hypothesis seems highly unlikely.
GOVERNMENT FORECASTS FOR EMPLOYMENT IN THE NON-FARM MARKET SECTOR Government forecasts clearly illustrate the different messages circulating with these figures, depending on whether they are presented yearon-year or as the annual average3.
In 2021, the most positive signal is from the year-on-year change in employment, whereas in 2022 it is the annual average increase. The government’s forecasts are summarised in Table 1: the most recent figures are from the April 2021 edition of the Stability Programme, while the others are part of the 2021 budget, published in October 2020. This table also illustrates how employment surprised favourably in 2020: year-on-year job losses were a little over half the size previously expected. The more limited employment fall in 2020 translates into a more limited rebound in 2021 on a year-on-year basis, in the most recent forecasts for the Stability Programme.
Since the negative carry-over was smaller, however, the 2021 annual average increase swung into positive territory, albeit slightly.
ESTIMATED IMPACT OF REFORMS ON EMPLOYMENTThe reason why we expect to see a mild rebound in employment in 2021 is because there will be fewer and less strong tailwinds than headwinds.
Among the support factors, there is of course the expected rebound in growth. There are also the specific positive effects of the youth employment plan, and more globally, of the France Relance recovery plan (+240,000 jobs by 2022 and +120,000 in the long term, according to government estimates presented in the Economic, Social and Financial Report as part of the 2021 budget).
There is also the impact of the reforms already underway (see chart 2). Yet caution is still needed given the high uncertainty over the vigour of the recovery, the impact of support measures and the job-rich content of the recovery.
However, Pôle emploi’s 2021 survey on companies’ labour need, released early May, sends an encouraging signal, with 2.72 million hiring plans, or 30,000 more than in 20194.
AFTER A LIMITED DECLINE, A MILD REBOUND?
The job content of growth is likely to be limited by corporate efforts to rebuild productivity gains and margins. Considering the deterioration in both variables due to the Covid-19 crisis, we can imagine that substantial efforts will be made to close the gap. The bigger the effort, however, the smaller the rebound in employment.
The job content of growth will also be limited by the return to work of furloughed or short-time workers. Job-retention schemes have proven to be an effective way to contain job destructions in the face of the massive recessionary shock. This effectiveness is good news, although it does raise a few questions. By limiting job destructions during the crisis, to what extent will it also limit job creations during the recovery phase? Is this trade-off preferable to the contrary, as is the case in the United States, where the labour market is left on its own and flexibility is allowed to run its course, resulting in more job losses followed by more job gains.
As to France, the answer – without much hesitation – is positive: job retention measures preserve human capital and are preferable, especially given the functioning and rigidity of the French labour market. According to the IMF (2021), regardless of the country, there can be no doubt about the effectiveness of job retention measures in buffering the shock of the Covid-19 crisis and its scars on employment5.
American laissez-faire also has its limits. Job creations since May 2020 have certainly been impressive (14 million new jobs), but the shock of the crisis was so big that there is still a substantial shortfall of jobs (in March 2021, employment was still down by 8 million jobs compared to pre-crisis levels, and only 63% of job losses had been recouped). And we are only speaking about the return to pre-crisis levels, and not to the pre-crisis trajectory. The US unemployment rate is still nearly 3 points above pre-crisis levels (see charts 1 and 2). The participation rate has also declined. In France, by comparison, there was still a deficit of about 300,000 payroll jobs in Q4 2020 compared to Q4 2019, 60% of job losses had been recovered, and the unemployment rate was 0.1 points lower than the pre-crisis level.
NON-FARM PAYROLL EMPLOYMENT IN FRANCE AND THE UNITED STATESTo buffer household income from the shock of the crisis, and lacking the same automatic stabilisers, the American fiscal response has been very different from the French one. In France, household income was preserved via employment, through the massive use of furlough or short-time work schemes. In the US, it was preserved through a big increase in unemployment benefits and through the direct distribution of stimulus checks6.
Although the amount of funds injected by the US Government is truly impressive, it nonetheless raises questions about the results. Did they live up to the amounts invested? Based on the first series of US and French measures, Cohen-Setton and Pisani-Ferry (2020) respond in the negative: in terms of their effectiveness, France’s package offered a bigger bang for the buck7.
Turning back to the headwinds, we must also take into account the lagged effect of the massive recessionary shock of 2020, the extra costs of the Covid-19 crisis, and the impact of the expected increase in the number of bankruptcies, one of the scars of the crisis. The OFCE (2020a) estimates that bankruptcies will result in 180,000 payroll job destructions in 20218, a figure that corroborates with the one proposed by Banerjee et al. (2020) for all of the developed countries. According to this study, a crisis has a much bigger impact on the labour market when it is accompanied by corporate bankruptcies similar in scope. Banerjee et al. also found that unbalanced recessions (i.e. in which corporate bankruptcies are concentrated more in certain sectors) tend to be longer and deeper, and to have a more negative impact on the labour market. The Covid-19 crisis, with its wide disparities in sector impact, fits this description.
The rebound in payroll employment is also likely to be hampered by the fact that the services that were hit hardest by the crisis (hotels & restaurants, leisure services, transport services) are also major employers (15% of total employment in 2019 for 10% of gross value added). Once the health restrictions have been lifted for good, we can certainly imagine that business and hiring will pick up rapidly. For this to happen, however, the number of corporate bankruptcies must remain contained (and/or new companies must rapidly replace the failing ones) and companies must be able to easily hire the necessary workers. This is far from certain.
BREAKDOWN OF PAYROLL EMPLOYMENT TRENDS BY FACTOR (‘000S, YEAR-ON-YEAR AT END OF YEAR)Lastly, there is also the more global question of whether certain business models need to be called into question in a post-crisis world marked by the challenges of climate change.
Crises are certainly catalysts for change – for creative destruction – which eventually has a net positive effect on employment. Yet Banerjee et al. point out that for that to happen, the reallocation of human capital and reskilling/ upskilling efforts must proceed rapidly and effectively. Although the French government has launched measures along these lines, their impact will only come into play over the long term. The year 2021 does not provide enough of a timeframe to judge whether these measures will be sufficient or effective. And in any case, we must be vigilant that they actually bear fruit.
To summarize, Table 2 lists most of the factors mentioned above and shows their expected impact on employment as estimated by the OFCE (2020b)10.
With virtually no net impact at year-end 2021, the message is not very optimistic.Yet these figures must be kept in perspective since the estimates date from late 2020 and do not include Q4 2020 data. The situation was not as bad as expected, with 284k payroll job losses compared to the year-earlier period (vs estimates of 790k). Even if the net impact is still zero, using a less negative starting point brings us into positive terri-tory in 2021, which is symbolically important. The estimated impact of job support measures in 2021 is probably too low as well.
UNEMPLOYMENT: HOW HIGH SHOULD IT BE EXPECTED TO RISE?
TWO NEW GAUGES TO ASSESS THE HEALTH OF THE FRENCH LABOUR MARKETAs to the unemployment rate, there are two sources of uncertainty concerning the size of its rise in 2021. There is uncertainty over the size of the rebound in employment but also in the labour force. The unemployment rate could rise sharply under the combined impact of a limited rebound in employment and a big upturn in the labour force. The labour force could swell as discouraged workers and those prevented from actively seeking work during the crisis – especially during the lockdowns – return to the labour market.
The ranks of the unemployed (as well as the number of corporate bankruptcies) are also likely to grow once emergency measures have been lifted, especially short-time work schemes.
There is reason to hope, however, that this increase will be limited, since the emergency measures will not be withdrawn until economic conditions are ripe. This should reduce the risk of a surge in unemployment (and bankruptcies). Moreover, support will be withdrawn gradually, and it will not be withdrawn entirely thanks to the introduction of a new long-term furlough scheme (see table and box below for a summary of the changes made to short-time work schemes and a summary of other measures to bolster the labour market).
These points can be illustrated with a few figures. In mid-April, we estimated that the unemployment rate would rise to 9.4% in Q4 2021, and to an average annual rate of 9.1%, which is a relatively big increase (+1.4 points compared to Q4 2020, +1.1 point compared to 2020). Our forecast is lower than the April 2021 consensus forecast (9.4%; with a forecast range of 8.4% to 10.7%11), but slightly higher than the Bank of France’s March 2021 forecast (average annual unemployment rate of 8.9%) and the IMF’s April 2021 forecast (expected average rise of 0.9 points).
In addition to the size of the increase in the jobless rate in 2021, there is also the question of its profile: will the jobless rate increase continuously over the course of the year or will it be hump-shaped? A bell-shaped profile seems to be preferable since it would signify that the unemployment rate had begun to decline by the end of the year. Yet it also depends on the shape of the hump, whether it is flatter or more acute, reflecting a lower or higher peak in the unemployment rate. This brings us back to the question of the size of the rise. For the moment, these questions cannot be answered12. In terms of probability, we would say that the most likely scenario is a continuous rise in the unemployment rate, that is steeper in the first half and that eases in the second half.
To conclude, mixed labour market prospects make 2021 a transition year: compared to 2020, there should be a definite improvement on the employment front but not in terms of the unemployment rate, while bigger improvements are expected on both fronts in 2022. To determine when the French labour market has returned to good health, we can use the two gauges retained by the French government in the unemployment insurance reform:
1/ the number of category A jobseekers must decline by at least 130,000 in 6 months;
2/ hiring reports for jobs lasting more than 1 month (excluding temporary work) must exceed a 4-month moving average of 2.7 million contracts. In chart 4, the two variables are expressed as the differential from their corresponding benchmark. The two spreads read in the same direction: a positive spread is a good sign and a negative one is bad.
Judging by the available data since 2006, the French government seems to have set the bar very high: the two criteria have never been met at the same time. The peak in October 2020 does not count. The positive signal is purely artificial and ephemeral because it is based on a favourable basis effect. The labour market situation was by no means healthy at the time, and was still crippled by the Covid-19 crisis. The threshold was almost reached in early 2020, prior to the Covid-19 shock. Without the crisis13, it is highly probable that the situation would have continued to improve, and the government’s target would have been reached. But everything changed with the onslaught of the Covid-19 crisis. By when can we now expect to see the French labour market return to good health? The year 2021 seems like an improbable horizon, although it could be envisioned in 2022.
