As with the economic performance of all the industrialized
countries, economic activity fell off sharply in the second quarter of 2020
across the Atlantic before rebounding just as sharply the following quarter. The
management of the crisis in the US is largely in the hands of the different States,
and the election of Joe Biden should not change this framework since he
declared on November 19 that he would not order a national lockdown. However,
the health situation is continuing to deteriorate, with more than 200,000 new Covid-19
cases per day on average since the beginning of December. As a result, many
States are adopting more restrictive prophylactic measures, although without returning
to a lockdown like the one in the Spring. This situation could dampen economic prospects
for the end of the year and also for the start of the mandate of the new
President elected in November. Above all, it makes it even more necessary to
implement a new recovery plan, which was delayed by the election.
As in the euro zone, recovery in the US kicked off as
soon as the lockdown was lifted. GDP grew by 7.4% in the third quarter after
falling by 9% in the previous quarter. Compared with the level of activity at
the end of 2019, the economic downturn amounted to 3.5 points, versus 4.4
points in the euro zone. The labour market situation also improved rapidly,
with the unemployment rate falling by 8 points, according to data from the Bureau
of Labor Statistics for November, from its April peak of 14.7%. These results
are the logical consequence of the lifting of restrictions but also of the large-scale
stimulus plans approved in March and April, which have massively absorbed the
loss of income for households and to a lesser extent for US companies (see here).
However, the upturn in consumption is still being dampened by some ongoing restrictions,
particularly in sectors with strong social interactions, where spending is
still nearly 25% lower than it was in the fourth quarter of 2019 (Figure 1). As for the consumption of goods, it has been much less
affected by the crisis and is down only 12% from its pre-crisis level for
durable goods and 4.4% for non-durable goods. Nevertheless, most of these
support measures have come to an end, and as of this writing the discussions
that began in late summer in Congress have not yet led to an agreement between
Republicans and Democrats. Despite the rebound, the health impact of the pandemic
and the economic consequences of the lockdown on the labour market require a discretionary
policy in a country where the automatic stabilizers are generally considered to
be weaker[1]. New support measures will be all the more
necessary as a further tightening of restrictions is looming and the recovery
seem to be running out of steam. The initial consumption figures for the month
of October point to a fall in the consumption of services, and employment also
stabilized in November, remaining well below its level at the end of 2019.
However, after the setback of the discussions in
Congress, it will now be necessary to wait until the first quarter of 2021 for
a new support plan to be approved and for a possible reorientation of US fiscal
policy after Joe Biden’s victory. In the Autumn, the Democrats proposed a 2
trillion dollar (9.5 GDP points) package, almost as much as the 2.4 trillion dollar
(10.6 GDP points) package adopted in March-April 2020[2]. The aid would, among other things, support the
purchasing power of the unemployed through an additional federal payment.
Although unemployment is much lower than in the second quarter, it remains
above its pre-crisis level and is now characterized by an increase in long-term
unemployment for which there is generally no compensation. In November, the
share of those who had been unemployed for at least 27 weeks was 37 per cent
(or 3.9 million people, Figure 2), and the median duration of unemployment
had risen from 9 weeks at the end of 2019 to almost 19 weeks in November 2020.
In addition, States whose tax revenues have decreased with the crisis could
benefit from a federal transfer, thereby avoiding spending cuts[3].
However, despite the end of the suspense over the
outcome of the presidential elections, the political and economic uncertainty
has not been completely resolved. Indeed, it will not be known until early
January whether the Democrats will also have a majority in Congress. They have
certainly kept the House of Representatives, but it will be necessary to wait
until the beginning of January for the Senate, with a ballot planned in Georgia
that will determine the political colour of the last two seats [4]. Both seats are now held by Republican senators.
However, Joe Biden won Georgia by 0.2 points against Donald Trump, the first
victory in the State for a Democratic candidate since 1992. With both State-wide
senatorial elections to be contested directly, the results are likely to be
close. If one of the Democratic
candidates is defeated, Joe Biden will be forced to contend with the
opposition. But, as Paul Krugman
points out, the Republicans are generally more inclined, once in opposition, to
promote austerity. This is reflected in the uncertainty indicators of Bloom,
Baker and Davies, whose economic policy uncertainty rose in November (Figure 3).
This uncertainty is certainly lower than in the Spring but remains higher than
that observed between 2016 and 2019. During this period, growth could weaken,
and then a strong recovery is likely to be followed by more subdued growth,
which will have repercussions on the labour market. Regardless of the outcome,
a plan will likely be approved in the first quarter of 2021, but its adoption
could take longer if it is conditional on an agreement between Republicans and
Democrats in Congress. However, this could be lengthy given the urgency of the
health and social crisis, and could plunge a significant proportion of the most
vulnerable into poverty.
[1] See for example Dolls, M., Fuest, C. &
Peichl, A., 2012, “Automatic stabilizers and economic crisis: US vs. Europe”, Journal of Public Economics,
96(3-4), pp. 279-294.
[2] By comparison, the
European programmes are weaker, ranging from 2.6 GDP points for France to 7.2
points for the UK.
[3] Note that the States generally have fiscal
rules limiting their capacity to run a deficit.
The sharp fall in activity and its brutal social consequences have led governments and central banks to enact ambitious support measures to cushion the shock, which resulted in an unprecedented global recession in the first half of 2020, as discussed in Policy Brief 78 . Faced with a health crisis that is unprecedented in contemporary history, requiring forced shutdowns to curb the spread of the virus, governments have taken urgent measures to prevent the onset of an uncontrolled crisis that could permanently alter the economic trajectory. Three main types of measures have been taken: some aim to maintain consumer purchasing power in the face of the shutdowns; others seek to preserve the production system by targeting business; and some are specific to the health sector. The quarterly national accounts, available at the end of the first half of the year, provide an update on the extent to which the disposable income of private agents has been preserved by fiscal policy at this stage of the Covid-19 crisis [2].
Fiscal policy has shot up Americans’ household
income and preserved Europeans’ income
In the major advanced economies, the Covid-19
crisis generated losses in primary income (before cash transfers) ranging from 81
billion pounds in the United Kingdom to 458 billion dollars in the United
States (Table 1). The initial income shock was thus larger in Spain and Italy –
6.5 and 6.7 GDP points respectively – and smaller in Germany (3.4 GDP points)
and the United States (2.1 GDP points).
Figure 1 breaks down the share of the primary income (PI) shock received by agents (first bar on the left for each country, labelled “PI”). In Spain and Italy, households suffered the majority of the losses, accounting for 54 percent and 60 percent, respectively, of the total income loss for the economy. In France and Germany, enterprises bore the lion’s share of the income loss (48%). In the United Kingdom and the United States, enterprises incurred losses of £50 billion and $275 billion, respectively, accounting for 62% and 60% of the total loss for the economy. General government (GG) experienced a smaller shock in all the countries, which is explained by the spontaneous changes in some of the automatic stabilizers, and by a relatively lower value added due to the restrictions on activity during lockdowns.
Turning to the breakdown in losses in disposable
income (DI), which takes into account cash transfers, social contributions, and
income tax, the story is rather different. The implementation of emergency
measures made it possible to absorb some of these losses, as illustrated by the
bar labelled “DI” in Figure 1. The introduction of short-time working
in European countries thus shifted the burden of wages from enterprises to the
government, thus preserving household incomes and avoiding the termination of job
contracts. Similarly, reductions in social contributions and tax on income and
corporate profits have shifted the cost of the crisis from private agents to
government. In the face of the unforeseeable shock, the State has thus played
the role of insurer of last resort of private agent income, although to
different extents in different countries. Thus, while Spain’s government absorbed
13.5 percent of the primary income shock, support measures raised this share to
59 percent, a higher level than that of Italy (55.3 percent) and France (54.3
percent) in terms of disposable income. In comparison, the measures taken by
the German government absorbed a higher share of the shock, amounting to 67
percent of the loss of disposable income, compared with 28 percent of the fall in
primary income.
In the United Kingdom, emergency measures absorbed
the entirety of the shock. While business and households suffered primary
income losses of £50 billion and £15 billion respectively, their disposable
income fell by only £4 billion and £2 billion. As for disposable income,
government absorbed 93.6 percent of the shock. The contrast is even more marked
in Germany and the United States, where measures overcompensated the initial
primary income shock, especially for households. The US figures are
particularly impressive. Over the six-month period, primary income fell by $192
billion, while household disposable income rose by $576 billion, due in
particular to the payment of a tax credit and an exceptional federal
unemployment benefit of $600 per week that was paid to the unemployed,
regardless of their initial income[3]. The various tax measures and subsidies to
business reduced the loss by $210 billion. The US government thus absorbed 237
per cent of the shock, reflecting the magnitude of the support measures taken
in March-April.
Job losses and uncertainty about the future may
hamper recovery across the Atlantic
As we have seen, fiscal policy has been mobilized
massively across the Atlantic. Even if at this stage the macroeconomic shock has
been weaker in the US than in the EU[4], the fiscal impulse is much larger. At the end of
the first half-year, total transfers to households exceeded the immediate shock
to their primary income. This has led to a 13% increase in the disposable
income of US households, at the same time as their primary income fell by 4% in
connection with job destruction. This situation is due in particular to a tax credit
paid to households and an additional lump-sum allowance of $600 per week paid
by the federal government to any person eligible for unemployment. Between Q4 of
2019 and Q2 of 2020, transfers to households leapt by 80%, now representing 31%
of disposable income compared with 19% in 2019.
This difference in crisis management is undoubtedly
explained by the weakness of the social safety net in the United States, which
effectively reduces the role of automatic stabilizers while also limiting the
ability of citizens with little or no health insurance coverage to meet health
care expenses in the event of a fall in income. The use of counter-cyclical
measures is thus of greater importance, which probably explains why the
stimulus packages are more extensive than they were during the 2008-2009 crisis
as well as why the measures provide direct, substantial support to household
income. Moreover, in the US, the federal government is responsible for this
stimulus, while in the EU, the bulk of the support plans come from the Member states.
The sharp rise in unemployment across the Atlantic
– which peaked at 14.7% in April – contrasts with the situation in Europe,
partly due to the differentiated strategy in economic policy. The United States carried out a positive, substantial
transfer of income to households to offset the fall in wages resulting from job
losses, which also helped to mitigate the shock on business margins.
Conversely, in the main European economies, contractual employment
relationships were maintained, but household incomes were not preserved quite
as much – they actually fell slightly, except in Germany. In the main European
economies, a decision was taken to use short-time working on a massive scale, while
in the United States the response was to send cheques directly and immediately
to households.
This situation, where income was propped up during
a period when consumption was curtailed by the closure of non-essential shops, led
to the accumulation of 76 billion euros in “Covid savings” in Germany
(8 GDI points), 62 billion in France (9 GDI points) and 38 billion in Spain and
Italy (10 and 6 GDI points respectively). In the United Kingdom and the United
States, “Covid savings” were even greater: £89 billion in the UK (12 GDI
points), while the sum reached $961 billion in the US (12 GDI points). How the
epidemic develops and how these savings are used will be the two keys
determining the extent of the rebound in activity starting in the second half
of 2020.
This is precisely the moment when differences in
approach can create divergences in economic trajectories. While it could be
said that up to now household situations have been better preserved across the
Atlantic, job contracts have been shredded. In this context, it may take some
time to get the workforce back into employment, hindering the rapid
redeployment of the production base. This could slow down the speed at which activity
returns to normal, helping to keep job losses up and limiting the restoration
of company balance sheets. Furthermore, negotiations between Democrats and
Republicans in Congress have hit the wall of the approaching November 3
elections. If the measures taken during the crisis are not – at least partially
– renewed, the situation of American households is likely to become more
critical, since weak US social safety nets will not be able to mitigate what
threatens to be a long-term shock. This may have second-round effects on
primary income and investment [5]. Following the elections, further measures are
likely to be taken, but the time lag could be long, especially if Joe Biden
wins, as he will have to wait until he takes office in January 2021. Continued
high uncertainty about the extent of the recovery – accentuated by political
uncertainty – may encourage American households to avoid spending “Covid
savings” in order to have “precautionary savings” to face a probable
long-term health, economic and social crisis.
Glossary
Primary income (PI): Primary income includes revenue directly related
to participation in the production process. The bulk of primary household
income consists of wages, salaries and property income.
Gross disposable income (GDI): Income available to agents to consume or invest,
after redistribution operations. This includes primary income plus social cash
benefits and minus social contributions and taxes paid.
[2] These results should be taken with a grain of
salt. While the quarterly national accounts are the most comprehensive,
consistent framework available, with data collected by official statistics
institutes, they are nevertheless provisional. These accounts are subject to
significant revisions that may significantly alter the final results when they
incorporate new data (company balance sheets, etc.); they are considered final
within two years.
[3] This allowance is in addition to that paid by
State-run unemployment insurance systems.
[4] The loss in 6-month GDP was 5% in the US,
compared with 8.3% in the EU.
[5] F. Buera, R. Fattal-Jaef, H. Hopenhayn, A.
Neumeyer, and J. Shin (2020), “The Economic Ripple Effects of COVID-19”, Working Paper.
What more could the central banks do to deal with the crisis?
The return of new lockdown measures in numerous countries
is expected to slow the pace of economic recovery and even lead to another
downturn in activity towards the end of the year. To address this risk,
governments are announcing new support measures that in some cases supplement
the stimulus plans enacted in the autumn. No additional monetary policy
measures have yet been announced. But with rates close to or at 0% and with a
massive bond purchase policy, one wonders whether the central banks still have any
manoeuvring room. In practice, they could continue QE programmes and increase
the volume of asset purchases. But other options are also conceivable, such as
monetizing the public debt.
With the Covid-19 crisis, the central banks – the
Federal Reserve, the Bank of England and the ECB – have resumed or amplified
their quantitative easing (QE) policy, to such an extent that some are viewing
this as a de facto monetization of debt. In a recent Policy
Brief, we argue that QE cannot
strictly be considered as the monetization of public debt, in particular
because the purchases of securities are not matched by the issuance of money
but by the issuance of excess reserves. These are distinct from the currency in
circulation in the economy, since they can be used only within the banking
system and are subject to an interest rate (the deposit facility rate in the
case of the euro zone), unlike currency in circulation.
Our analysis therefore makes it possible to look
again at the characteristics of QE and to specify the conditions for monetizing
debt. It should result in (1) a saving of interest paid by the government, (2) the
creation of money, (3) being permanent (or sustainable), and (4) reflect an
implicit change in the objective of the central banks or their inflation
target. The implementation of such a strategy is therefore an option available
to central banks and would allow the financing of expansionary fiscal policies.
The government, in return for a package of fiscal measures – transfers to
households or health care spending, support for businesses – would issue a
zero-coupon perpetual bond, purchased by commercial banks, which would credit
the account of the agents targeted by the support measures. The debt would have
no repayment or interest payment obligations and would then be acquired by the
central bank and retained on its balance sheet.
Monetization would probably be more effective than QE
in stabilizing nominal growth. It would reduce the risk to financial stability caused
by QE, whose effect depends on its transmission to asset prices, which could
create asset-price bubbles or induce private agents to take on excessive debt.
Monetization has often been put off because of fears that it would lead to
higher inflation. In the current environment, expansionary fiscal policy is
needed to sustain activity and to prepare for recovery once the pandemic is
under control. A pick-up in the pace of inflation would also satisfy the central
banks, and insufficient demand should greatly reduce the risk of an out-of-control
inflationary spiral. Monetization requires stronger coordination with fiscal
policy, which makes it more difficult to implement in the euro area.
Europe’s recovery plan: Watch out for inconsistency!
On 27 May, the European Commission proposed the
creation of a new financial instrument, Next Generation EU,
endowed with 750 billion euros. The plan rests on several pillars, and will notably
be accompanied by a new scheme to promote the revival of activity in the
countries hit hardest by the coronavirus crisis. It comes on top of the
Pandemic Crisis Support adopted by the European Council in April 2020. A new
programme called the Recovery and Resilience Facility will have firepower of 560
billion euros, roughly the same amount as the Pandemic Crisis Support. The
Recovery and Resilience Facility stands out, however, for two reasons: first,
by the fact that part of its budget will go to grants rather than loans; and
second, by its much longer time horizon.
The Pandemic Crisis Support (and the complementary
tools adopted at that time, see Creel, Ragot & Saraceno, 2020) consists exclusively of loans, and the net gains that
the Member States could draw from them are by definition low: European loans
allow a reduction in interest charges for States subject to high interest rates
on the markets. The gain for Italy, which was hurt badly by the coronavirus
crisis, is in the range of 0.04 to 0.08% of its GDP (this is not a typo!).
Under the Recovery and Resilience Facility, the euro
zone Member States would share 193 billion euros in loans and 241 billion euros
in grants, or in total 78% of the amounts allocated (the rest will go to EU states
that are not euro zone members). The loans will generate small net gains for Member
States (savings on the infamous interest rate spreads), while the grants will lead
to larger gains, since they will not be subject to repayment, other than via higher
contributions between 2028 and 2058 to the European budget (if the EU’s own funds
have not been created or increased by then). In the short term, in any case,
the grants received represent net gains for the beneficiaries: they will
neither need to issue debt nor pay interest charges on such debt.
Expressed as a percentage of 2019 GDP, the net
gains from grants are far from negligible (Table 1)[2]: 9 GDP points for Greece, 6 for Portugal, 5 for
Spain and 3.5 for Italy. This will be even more significant given the expected
fall in GDP in 2020. The determination of the Commission is therefore clear.
Despite all this, these grants are not intended to
be used in the short term. The European Commission purportedly wanted the
allocated amounts to be spent as quickly as possible, in 2021, 2022 and in any
case before 2024. This is what it calls “front-loading”: do not put
off till the morrow what can be done today. Except that the key to the
distribution of the grant expenditures over time is somewhat in contradiction
with this principle (Table 2). The grant commitments would be concentrated in
2021 and 2022, but the actual disbursals are planned for later: less than a
quarter by 2023, half in 2023 and 2024, and the remainder after that. This kind
of gap is frequent: it takes a little time to design an investment project and
to ensure that it complies with the European Commission’s digital ambitions and
low-carbon economy.
As a result, the grants to the Member States will
take a little time to actually be disbursed (Table 3), and the countries facing
the greatest difficulties will have to be resilient before receiving the stimulus
and… resilience funds. This seems contradictory. It will take until 2022 in
Greece and Portugal and 2023 in Spain and Italy to actually collect around 1
GDP point apiece. This corresponds to 3 billion euros for Greece, 2 billion for
Portugal, and 14 for Spain and Italy, respectively. By way of comparison,
Germany, France and the Netherlands will by then receive 5, 7 and 1 billion
euros, respectively, i.e. between 0.2 and 0.3 percent of their GDPs.
One can imagine the cries of outrage from the representatives of the frugal countries (Austria, Denmark, the Netherlands, Sweden) that these immense outgoings reward countries that are not virtuous. They should be reassured: this is no boondoggle!
[1] This text appeared in the 23 May 2020 edition
of Les Echos, without the tables.
[2] The rule for the distribution of transfers
between countries appears in the document COM (2020) 408 final/3 of 2 June
2020. For each country it depends on the size of its population, on the inverse
of GDP per capita compared to the EU-27 average, and on the difference between its
5-year unemployment rate and the EU-27 average. In order to avoid an excessive
concentration of grants to a few countries, ad hoc limits are imposed based on
these three criteria. Germany will for example receive 7% of the transfers,
France 10%, and Spain and Italy 20%, respectively.
Sweden and Covid-19: No lockdown doesn’t mean no recession
Since the Covid-19 pandemic’s
arrival on the old continent, a number of countries have taken strict measures
to limit outbreaks of contamination. Italy, Spain, France and the United
Kingdom belatedly stood out with especially strict measures, including lockdowns
of the population not working in key sectors. Sweden, in contrast, has
distinguished itself by the absence of any lockdown. While public events have
been banned, as in the other major European countries, there were no
administrative orders to close shops or to impose legal constraints on domestic
travel[1].
Given the
multiplicity of measures and their qualitative nature, it is difficult to break
down all the decisions taken, and in particular to express their intensity.
Researchers at the University of Oxford and the Blavatnik School of Government
have nevertheless built an indicator to measure the severity of government
responses[2]. This indicator clearly shows Sweden’s specific
situation with respect to the rest of Europe (Figure 1).
The mobility data supplied
by Apple Mobility provides a complementary picture of the severity of
containment measures across countries. At the time of the toughest lockdowns, automobile
mobility was down by 89% in Spain, 87% in Italy, 85% in France and 76% in the
United Kingdom. The decline was less severe in Germany and the United States
(about 60% in both countries). Sweden ultimately saw its traffic reduced by
“only” 23%. While these data should be taken with a grain of salt,
they also give a clear signal about the timing and scale of the lockdowns in
different countries, once again pointing to a Swedish exception.
During the first half
of May, the various European countries began to gradually ease the measures
taken to combat the spread of the Covid-19 epidemic.
Sweden’s
GDP resists in Q1
In our assessment of
the impact of lockdowns on the global economy, we highlighted the correlation between the fall in
GDP observed in the first quarter and the severity of the measures put in place
to combat Covid-19. In this context, Sweden (in red in Figure 2) fares
significantly better than the OECD member countries (green bar), and especially
the rest of the European Union (purple bar). Although this is a first estimate,
GDP has not only held up better than elsewhere, but has even stabilized (‑0.1%).
Only a few emerging economies, which were not affected by the pandemic at the
beginning of the year (Chile, India, Turkey and Russia), and Ireland, which
benefited from exceptional factors, performed better in the first quarter [3].
The relative
resilience of Sweden’s GDP in the first quarter seems to suggest that the
country might have found a different trade-off between epidemiological and
economic objectives compared to other countries[4]. However, this aggregate figure masks important
developments that need to be kept in mind. In the first quarter,
the stabilisation of Swedish GDP was due to the positive contribution made by foreign
trade (up 1.7 GDP points) to a rise in exports (up 3.4% in volume terms),
particularly in January, before any health measures were taken.
In the first quarter,
Swedish domestic demand pulled activity downwards (by ‑0.8 GDP points due to household
consumption and -0.2 GDP points due to investment), as in the rest of the EU. The
shock to domestic demand was of course more moderate than in the euro area,
where consumption contributed negatively to GDP by 2.5 points and investment by
0.9 points. Nevertheless, the physical distancing guidelines issued in Sweden must
have had a significant impact during the first quarter.
In a
troubled global context, Sweden will not be able to escape a recession
If we assume that the
avoidance of a lockdown and the relatively limited administrative closures (confined
to public events) did not give rise to any significant shock to domestic demand
– which seems optimistic in view of the first quarter data – Sweden will
nevertheless be hit hard by the shock to international trade[5].
According
to our calculations, based on the entry-exit tables from the World Input-Output
Database (WIOD)[6] and our estimates related to the
lockdown shocks in Policy Brief 69, value added is expected to fall by
8.5 points in Sweden in April due to the containment measures taken in the rest
of the world. The shock will hit its industry especially hard, more or less in
line with what we estimate globally (-19% and 21%, respectively).
Unsurprisingly, the refining industry (-32%), the manufacture of
transport equipment (-30%) and capital goods (-20%), and the other
manufacturing industries sector (-20%) will be hit hardest by the collapse
of global activity. Since a significant share of output is intended for use by
foreign industry, the worldwide containment measures will lead to a reduction
of almost 15 points in Swedish output in April (Figure 3). The same holds for commercial
services: exposure to global production chains is hurting transport and warehousing
(-15%) and the business services sector (-11%). Ultimately, the containment
measures will have an impact mainly through their effect on intra-branch trade.
The
weakness of Swedish manufacturing, weighed down by international trade, seems
to be confirmed by the first hard data available. According to the Swedish Statistical Office, exports fell by 17% year-on-year, a
figure comparable to the decline in world trade as measured by the CPB for the
same month (-16% by volume). Given this situation, manufacturing output will be
17% lower in April than a year earlier.
What
could be said about domestic demand in Q2?
In
a context of widespread uncertainty, domestic demand may continue to suffer.
Indeed, Swedish households can legitimately question the consequences of the
shock for jobs – mainly in industry – described above. On the other hand, fear
of the epidemic could deter consumers from making certain purchases involving
strong social interactions, even in the absence of legal constraints. What do
Swedish data from the beginning of Q2 tell us about Swedish domestic demand?
In
Sweden, consumer spending fell in March (-5% year-on-year). Note that the
country’s precautionary guidelines and physical distancing measures were
introduced on 10 March. The fall steepened in April, after the measures had in
force for a full month (-10% year-on-year). The measures in place hit purchases
of clothing (-37%), transport (-29%), hotels and catering (-29%) and leisure
(-11%). While the data remain patchy, May’s retail sales, an indicator that
does not cover the entire consumer sector, suggest that sales were still in a
dire state in clothing stores (-32%). In addition, new vehicle registrations
continued to fall in May (-15% month-on-month and -50% year-on-year). Pending
more recent data on activity in the rest of the economy, the volume of hours
worked[7] in May remains very low in hotels and
catering (-50%), and in household services and culture (-18%), suggesting that
significant and long-lasting losses to business can be expected.
On
the positive side, the data show a trend towards the normalization of household
purchases in May for certain consumer items. As in other European countries,
the recovery was particularly strong in household goods, where retail sales
returned to their pre-Covid level, and in sporting goods, while food
consumption remained buoyant.
Ultimately,
the health precautions taken by Sweden since the onset of containment measures are
akin to those implemented in the rest of Europe since the gradual easing of the
lockdowns. While the shocks to the consumption of certain items are less severe
than those observed in France, it is noticeable that, in the context of the
epidemic, some consumer goods could be severely affected even in the absence of
administrative closures. In addition to the recessionary impact imported from
the rest of the world, Sweden will also suffer due to domestic demand, which is
expected to remain limited particularly in certain sectors. The Swedish case
suggests that clothing, automobile, hotel and catering, and household services
and culture could suffer a lasting shock even in the absence of compulsory measures.
According to data available in May, this shock could reduce household
consumption by 8 percentage points, which represents 3 GDP points. How lasting the
shock is will depend on the way the epidemic develops in Sweden and in the rest
of the world.
[1] The Swedish institutional framework
helps to explain in part this differentiated response, which focuses more on
individual responsibility than on coercion (see https://voxeu.org/article/sweden-s-constitution-decides-its-exceptional-covid-19-policy). The country’s low population density
could also help explain the difference in behaviour vis-à-vis the rest of
Europe but not in relation to its Scandinavian neighbours.
[2] This indicator attempts to synthesize
the containment measures adopted according to two types of criteria: first, the
severity of the restriction for each measure taken (closure of schools and of businesses,
limitation of gatherings, cancellation of public events, confinement to the
home, closure of public transport, restrictions on domestic and international
travel) and second, whether a country’s measures are local or more generalized.
For a discussion of the indicator see Policy brief 69.
[3]Booming exports in March 2020 (up 39% in value) driven by strong
demand for pharmaceuticals and IT offset the fall in Ireland’s domestic demand during
the first quarter.
[4] This post on the OFCE blog does not
focus on the effectiveness of Swedish measures with regard to containing the
epidemic. Mortality from Covid-19 is higher in Sweden than in its neighbours (Norway,
Finland, Denmark), suggesting that it has run more epidemiological risks. This is
provoking a debate that goes well beyond the purpose of this post, but which does
deserve to be raised.
[5] International trade may actually impact
growth more than expected due to constraints on international tourism. In 2018,
Sweden actually ran a negative tourism deficit of 0.6% of GDP (source: OECD
Tourism Statistics Database), which could have an effect on domestic
activity if travel remains limited, especially during the summer.
[6] Timmer, M. P., Dietzenbacher, E., Los, B.,
Stehrer, R. and de Vries, G. J. (2015), “An Illustrated User Guide to the World
Input–Output Database: The Case of Global Automotive Production”, Review of International Economics., 23: 575–605
[7] In May, the volume of hours worked was
down 8% year-on-year (after -15%). The recovery in hours worked in May was due mainly
to manufacturing and construction. The recovery was less pronounced or even non-existent
in business services.
Germany on the slippery slope of the research tax credit
After years of
hesitation, the German parliament has just introduced a tax scheme to promote
investment in R&D. The decision precedes the Covid-19 crisis, but it may
well be heaven-sent for German business.
What factors motivated
Germany to take such a decision, four decades after the United States and
France, when it is among the world’s leading investors, in terms of both R&D
and innovation? Is this yet another instrument to boost its competitiveness?
And what will be the repercussions on R&D spending in France?
The German tax
incentive, which came into force in January 2020, offers companies a tax credit
equal to 25% of the declared R&D expenditure. The base is narrower than for
France’s research tax credit (CIR), since in Germany only wages are taken into
account (including employer social security contributions).[1] The 25% rate is, however, close to the French rate
(30%). A company’s eligible expenses are capped at two million euros; and the
tax credit for each firm will be limited to 500,000 euros (subcontracting is
subject to slightly different treatment). When a group has several subsidiaries
benefiting from the system, as part of a joint research programme, the total eligible
expenses are capped at 15 million euros (for a tax credit of 3.75
million).
By way of comparison,
among French companies who carry out R&D, SMEs receive an average of
131,000 euros for the CIR credit, mid-caps [fewer than 5,000 employees] 742,000
euros, and large corporations 5.6 million, according to the MESRI’s
figures. The highest amounts exceed 30 million euros (with few companies in
this category), but do not go much higher, because the CIR rate falls from 30%
to 5% of eligible R&D expenditure beyond the base threshold of 100 million
euros. Estimates of the annual loss in taxation for Germany (before taking into
account the macroeconomic effects) could amount to as much as five billion
euros. This is 80% of the French CIR credit, and on the same level as the
R&D tax incentives in the United Kingdom. Without the cap, the scheme would
cost the German federal government around 9 billion euros.[2]
The characteristics
of the scheme and the high level of German private R&D raise questions
about the Parliament’s real motivations. Indeed, one could wonder why it did
not opt for an “incremental” system, that is, base itself on the increase in
eligible R&D expenditure, as in the United States, or in France until 2003.
Admittedly, an incremental system would not support firms whose R&D is stagnating
or falling (in which case direct aid is more effective), but it avoids the
windfall effects of France’s CIR credit (Salies, 2017).
The cap limits, but does not eliminate, these effects.
The level of private
R&D spending is significantly higher in Germany than in any other EU Member
State (62.2 billion euros, excluding direct grants). France is far behind (27.5
billion euros), followed by Italy and Sweden (respectively 12.8 and 9.6
billion). A comparable ranking is obtained, for Germany, France and Italy, if
we measure the R&D effort (expenditure relative to GDP; Figure 1).
Germany is at almost the same level as Sweden (resp. 1.92 and 2.01 points).
Next come Denmark, Belgium, Austria and Finland. France is in 7th position with
1.44 points and Italy 13th with 0.71 point. Private research in Germany (excluding
subsidies) is only 0.08 GDP points below the 2% threshold set at the Barcelona
European Council in 2002 (the “Lisbon strategy”), which Sweden alone has
achieved. If subsidies are included, the private sector exceeds this threshold.
Since 2017, Germany’s domestic expenditure on R&D (private and public) has
also exceeded the 3% threshold. The argument advanced in 2009 by Spengel and Grittmann from ZEW that a tax incentive would allow German companies
to overcome private underinvestment in R&D is therefore not convincing, at
least from a European perspective.
At the global level,
three countries are of course doing better than Germany: the United States,
China and Japan, where the private sector spends 1.6 euros for every euro spent
by Germany. However, if the motivation of Germany’s Parliament for introducing
a tax incentive was to catch up with these countries, it would not have done so
only 40 years after the United States!
The introduction of a
tax incentive for R&D is less surprising if we consider changes in the
R&D effort. We have calculated the average growth rate of the R&D
effort for the 27 current Member States plus the United Kingdom, Norway and
Iceland over the period 2002-2017 (Figure 2).
The curve through the
cloud (logarithmic adjustment) reveals an almost inverse relationship between
the rate and the effort in 2002, suggesting a convergence of R&D efforts.
Obviously, many countries are in a period of catch-up with respect to investing
in research. Most of them are small, but the whole is significant. For example,
in 2017 countries where the R&D effort grew at a rate at least equal to Germany’s
(1.52%) spent 82.8 billion euros (subsidies included), or 1.2 times Germany’s
expenditure (68.7 billion).[3] The R&D effort of these countries amounted to
0.8 point of GDP in 2017.[4]
Could the German CIR credit
thus be a response to the slowdown in the country’s spending on R&D?
R&D expenditure behaves like other capital expenditure, i.e. it slows as
the level rises. Furthermore, the more countries have a high level of domestic spending
on R&D, the more they invest in R&D abroad. This results from the fact
that R&D expenditure is mainly by large corporations and multinationals; we
could cite, for example, Alphabet, Volkswagen and Sanofi, which in 2019 spent, respectively,
18.3 billion, 13.6 billion and 5.9 billion euros on R&D according to
figures from the EU
Industrial R&D Scoreboard. It is notable that the big multinationals open
R&D centres abroad to get closer to their export markets, as well as for
the bargaining power that these investments provide vis-à-vis local governments
(see the report by UNCTAD WIR, 2005). All the major pharmaceutical firms (Pfizer,
GlaxoSmithKline, AstraZeneca, Sanofi-Aventis, Novartis, Eli Lilly) have
established clinical research laboratories in India. Even France’s power supply
firm EDF has an R&D centre in Beijing, dedicated to networks, renewable
energies and the sustainable city. While this does not necessarily amount to substitution
with domestic R&D, it does indicate that there is a kind of plateau in a
given country for a company’s R&D expenditure. The German measure is
probably motivated by global competition to attract new R&D centres. This
is also the stated objective of France’s CIR credit.
Does the enactment of
a “German CIR” credit in favour of R&D bode well for France’s
competitiveness? Germany has a comparative advantage in the manufacturing
sector, which invests heavily in R&D. The new German tax scheme will
reinforce this advantage, without any risk of European litigation, since
R&D support falls under the exemptions to the European Commission’s control
system on state aid. France’s comparative advantage tends to be situated in
services. France’s R&D effort in services is more intense than in Germany:
0.28% of GDP in Germany and 0.67% in France. However, France stands out for
providing less public support for R&D investment by service companies. In
2015, public funding’s share of private research in services was 4% in France,
compared to 11% in Germany, according to an INSEE study.
The “German CIR” will only increase the relative price of French private
research in services in comparison with German research. However, the R&D content
of services determines the price, since it determines their technological
content. The German tax advantage will therefore accentuate the cost advantage
of the technological services which are themselves incorporated into
manufacturing value added. So this will in turn increase the cost advantage of
German manufacturers.
In addition, the
price of R&D is increasingly determined by personnel costs, whose share in
R&D has tended to rise in Italy and France and slightly too in Germany.
This share was roughly equal in the latter two countries in 2017: 61.8% in
Germany, and 59.7% in France.[5] Relative changes in researchers’ salaries will
have an impact on the difference in the amount of the tax credit between France
and Germany. As noted, the new scheme introduced across the Rhine is based only
on the costs of personnel. It could thus be conceptualized as a credit like
France’s Competitiveness and Employment Tax Credit (CICE) targeted at high-skilled
workers in the research sector (referring to the CICE credit before it transforms
into a reduction in employer social security contributions).
This is the reason
why we think that Germany has rather wanted to pursue its policy of lowering
corporate taxes. This was one of the motivations for France’s CIR reform in
2008, which “[can] be viewed as [fiscal] compensation for lower corporate
tax rates in other countries” (Lentile and Mairesse, 2009).
The median tax rate in the OECD applied to large corporations has fallen
continuously since 1995 (13 points over the period 1995-2018), from 35% to 22%.
However, the German rate, which has fluctuated between 29 and 30% since 2008,
is close to the French rate (around 32% in 2020; EC, 2020).
The opposition that could exist in the realm of “tax philosophy”,
between a French system based on a high rate and numerous provisions for
exemptions, and a German system based on a broad base and low rates, is not as strong
now that Germany has set up its own “CIR” credit.
This new incentive is
expected to enhance Germany’s attractiveness for R&D activities, which has
deteriorated somewhat (EY, 2020;
see also CNEPI, 2019).
Since 2011, the top three countries welcoming the most R&D centre projects were
the United Kingdom, followed by Germany and France. Since 2018, France has
hosted more projects than Germany (1197 against 971 in 2019), relegating
Germany to third place (this had already transpired in 2009, during the
financial crisis). The new tax credit should influence the trade-off of foreign
companies that are hesitating between France and Germany about where to set up.
It should also attract French companies to Germany, in the same way that a
significant share of private R&D activities carried out in France come from
foreign companies: 21% in 2015, for the percentage of expenditure as well as
the percentage of employed researchers (see Salies, 2020).
In accordance with European law, French companies established across the Rhine,
and liable for the “Körperschaftsteuer” (German corporate tax),
should be able to benefit from this niche.
Finally, private and
public R&D entities located in France should be able to benefit from the
tax incentive introduced in Germany, via subcontracting. But this will be only of
marginal benefit, for two reasons: the tradition of the German
“Mittelstand” has a culture favouring local networks, and the base
for outsourced activities is capped (as with France’s CIR credit). French
subcontractors will probably be able to benefit from authorizations, in the
same way as France’s research ministry, the MESRI, issues authorizations in Germany. Since 2009, Germany has recovered 6%
of the subcontracting approvals granted by the MESRI, the United Kingdom 4%,
etc. The majority of authorizations are granted to companies located in France
(75%).
Whatever the reasons
that motivated the German Parliament to introduce a tax incentive in favour of
R&D expenditure, it is certain that France has no interest in retiring its
own scheme. This does not mean France shouldn’t reform the CIR credit, as the
leverage effects are not as strong as expected; aid (direct and indirect), in
GDP points, has increased on average by 5.7% per year since 2000, whereas
R&D, also in GDP points, has increased only by 0.73% per year. The weak leverage
effect may have been the factor that for a long time discouraged Germany
from introducing a tax break to boost R&D.
In this period of
searching for ways to support business, it goes without saying that the
research tax credit will remain unchanged in France and could see the base for
the scheme expanded in Germany (in particular to help car manufacturers who
have been refused a plan for direct support).
It is nonetheless
regrettable that one of the reasons for Germany’s new scheme is probably to be
found in the inability of the Member States to advance the European Common
Corporate Consolidated Tax Base (CCCTB) directive, which provides for
harmonized R&D taxation for large firms by deducting R&D expenditure
from the tax base on corporate profits. The German CIR may well be in
competition with the French CIR, leading to transfers of R&D (by multinationals)
from one State to another. The net increase in R&D spending by European
companies remains to be estimated. Unless this spending increases, German
policy could be viewed as yet one more uncooperative tax policy coming at a
time when Europe is looking for common tax revenue.
[1]. The French CIR credit
includes, in addition to personnel costs, costs for the acquisition of patents,
standardization, allocations relating to the depreciation of buildings used for
research, etc.
[2]. Based on a private R&D expenditure of 62
billion euros in 2017 (direct aid excluded), we find 0.25 (the rate of the tax
credit), 0.6 (the share of salaries in R&D), yielding a credit of 9.3
billion euros.
[3]. The Netherlands, the United Kingdom, Slovenia,
Slovakia, Belgium, Latvia, Italy, Romania, Austria, Lithuania, Portugal,
Hungary, Estonia, Cyprus, Greece, Bulgaria, Poland and Malta.
[4]. The GDP of these countries (at market prices in
2017) is 2.5 times that of Germany.
[5] The increase in France and in Italy was +7 and +20
points respectively over the period 2000-2017.
How to spend it: A proposal for a European Covid-19 recovery programme
The Recovery Fund recently proposed by the EU Commission marks a sea-change in
European integration. Yet it will not
be enough to meet the challenges Europe faces. There has been much
public debate about financing, but little about the sort of concrete projects
that the EU should be putting public money into. We propose in Policy
Brief n°72 a 10-year, €2tn investment programme focusing on public health,
transport infrastructure and energy/decarbonisation.
The investment programme consists of two pillars. In a national
pillar Member States – broadly as in the Commission proposal – would be
allocated €500bn. Resources should be focused on the hardest-hit countries and
front-loaded: we suggest over a three-year horizon.
The bulk of
the money – €1.5tn – would be devoted to finance genuinely European projects, where there is an EU value added. We
describe a series of flagship initiatives that the EU could launch in the
fields of public health, transport infrastructure and energy/decarbonisation.
We call for
a strengthened EU public health agency
that invests in health-staff skills and then facilitates their flexible
deployment in emergencies, and is tasked with ensuring supplies of vital
medicines (Health4EU).
We present
costed proposals for two ambitious transport initiatives: a dedicated European
high-speed rail network, the Ultra-Rapid-Train,
with four-routes cutting travel times between EU capitals and regions, and,
alternatively, an integrated European
Silk Road initiative that combines transport modes on the Chinese model.
In the area
of energy/decarbonisation we seek to “electrify”
the Green Deal. We call for funding to accelerate the realisation of a
smart and integrated electricity grid for 100%-renewable energy transmission (e-highway), support for complementary
battery and green-hydrogen projects, and a programme, modelled on the SURE
initiative, to co-finance member-state decarbonisation and Just Transition
policies.
The crisis
induced by the pandemic, coming as it does on top of the financial and euro
crises, poses a huge challenge. The response needs to take account of the
longer-run structural challenges, and above all that of climate change. The
European Union should rise to these challenges in the reform of an ambitious medium-run recovery programme,
appropriately financed. An outline of such a programme is set out here
by way of illustration, but many permutations and options are available to
policymakers.
[1] Andrew Watt: Macroeconomic Policy Institute (IMK),
Düsseldorf; email Andrew-watt@boeckler.de.
Jérôme Creel, Francesco Saraceno: OFCE, Paris. Mario Holzner: wiiw Wien. Jérôme
Wittwer: University of Bordeaux.
The COVID-19 crisis and the US labour market: Rising inequality and precariousness in perspective
In the United States as in France, the
COVID-19 crisis has led to numerous measures restricting economic activities intended
to limit the spread of the virus. The result will be a fall in GDP, which is already
showing up in figures for the first quarter of 2020, and which will be much steeper
in the second quarter. In a country noted for its weak employment protection,
this unprecedented recession is quickly having repercussions on the labour
market, as reflected in the rise in the unemployment rate from a low point of 3.5%
in February to 14.7% in April, a level not seen since 1948. As Bruno
Ducoudré and Pierre Madec have recently demonstrated in the case of France,
the current crisis in the United States should also result in heightened inequalities
and insecurity. And the shock will be all the greater in the US since the
social safety net is less extensive there.
In the United States, the Covid-19 restrictions
were set not at the Federal level but by the various States at differing times.
The vast majority of States did decide however to close schools and
non-essential businesses and to encourage people to stay home. The lockdown was
thus imposed by California on March 19, followed by Illinois on March 21 and
New York State on March 22, but South Carolina didn’t follow until April 6.
North Dakota, South Dakota, Arkansas, Iowa and Nebraska have taken no action,
and three other States – Oklahoma, Utah and Wyoming – applied measures only in certain
counties, and not State-wide. However, by early April a large part of the
country had been locked down, with a varying degree of strictness, affecting between
92% and 97% of the population[1].
Which employees have been hit hardest by the crisis?
According to a survey by the US Bureau of Labor Statistics, almost 25%
of employees worked from home in 2017-2018. However, some employees said they
could have stayed at home to work but did not necessarily do so during the
reporting period. With the COVID-19 crisis and the incentives to modify the
organization of work, we can therefore consider that almost 29% of employees
could stay at home during the lockdown [2].
Furthermore, as the survey
carried out for France highlights, the implementation of teleworking is more
widespread among employees in management jobs and commercial or financial
activities. In 2017-2018, 60% of these people could have managed to work from home.
In contrast, fewer than 10% of workers in agriculture, construction, manufacturing
or transport services would have been able to telework during the crisis. Not surprisingly,
the survey also shows that the employees able to telework are also those at the
top of the wage distribution. For the top quartile, 61.5% of employees could
work at home compared with fewer than 10% for employees in the bottom quartile.
Mirroring these
elements, a more recent study analyzed which jobs would be most affected by the
lockdowns and in particular by the closure of non-essential businesses [3]. Six sectors are particularly exposed.
Logically enough, these include bars and restaurants, transport and travel,
entertainment, personal services, the retail trade and some manufacturing
industries. Based on employment data for the year 2019, these sectors represent
20.4% of total employment. With more than 12 million jobs, the bar and
restaurant sector is being hit hardest. This survey also shows that the most
exposed employees generally receive below-average pay. They are particularly
concentrated in the two lowest wage deciles. For example, the wage bill for bar
and restaurant workers represents barely 3% of the total wage bill but more
than 8% of employment. These people usually work in companies with fewer than
10 employees. This dimension is all the greater in the United States since
access to health insurance is often linked to the employer, whose obligations for
insurance provision depend on how many employees they have. Finally, by
crossing the distribution by sector and geography, it appears that Nevada,
Hawaii and to a lesser extent Florida (23.7%) concentrate a larger share of these
sectors, and therefore of the exposed jobs [4]. Conversely, Nebraska, Iowa and Arkansas
are among the States where these sectors account for a smaller share of
employment [5]. These three States have also not adopted lockdown
measures and should therefore be relatively spared from the rise in unemployment.
Unemployment statistics for the months of
March and April
confirm this outlook. In one year, the unemployment rate increased by 4.8
points for those in management jobs or commercial or financial activities,
while, over the same period, the rate rose by 23 points for service jobs and
almost 15 points for employees in production. The geographic disparities are
also significant. In California and Illinois, the first States to implement a
lockdown, the unemployment rate rose 11.3 and 12.2 points, respectively, in one
year. Conversely, the States that have not enacted lockdown measures are among
those where the unemployment rate has risen the least over the year. The
increase reached 5.2 points for Nebraska, 6.7 points for Arkansas and 7.5
points for Iowa, for example.
The structure of employment is, however, a
key factor determining the variation in unemployment. Despite fairly close starting
dates for the lockdowns in Connecticut and Michigan, the unemployment rate rose
only 4.2 points in the former versus over 18 points in industrial Michigan. The
statistics also confirm the exposure to the shock of Nevada and Hawaii, which
recorded the two largest increases: 24.2 and 19.6 points respectively, while
Minnesota, with a very low exposure, saw its unemployment rate rise by only 4.9
points, one of the smallest variations since April 2019. Likewise, the impact
has been relatively softer in the District of Columbia, where the unemployment
rate rose by 5.5 points.
Health under threat?
The deteriorating state of the labour
market will be accompanied by a deterioration in living conditions for millions
of Americans, especially if the end of the lockdowns is not synonymous with a
rapid rebound in activity, as Jerome Powell, Chairman of the Federal Reserve,
now fears. This would result in increased poverty for households that have lost
their jobs. Previous analyses indicate that workers at the bottom of the
distribution will be the most exposed, especially since, despite the measures taken to
extend unemployment insurance, the duration of benefits remains overall
shorter in the United States. To deal with the crisis, the Federal government
has spent USD 268 billion (or 1.3 percentage points of GDP) on unemployment
insurance to extend the duration and amount of compensation. This is in
addition to the tax credit of up to USD 1,200 for households without children [6].
The government has thus chosen to support incomes temporarily, but unlike the
partial unemployment schemes in force in France and in many other European
countries, it has not protected jobs [7].
The flexibility of the US labour market could, however, prove more advantageous
in so far as the recovery is rapid and differs depending on the sector.
Employees actually do not lose much of their skills and can more easily find a
job in another business sector. But a protracted crisis associated with persistently
higher unemployment would greatly increase poverty.
In addition, access to health insurance is
also often linked to employment. Indeed, 66% of insured Americans are covered
by their employer, who is obliged to offer health insurance in companies with
more than 50 employees. The corollary is that many workers risk losing their
health coverage at the same time as their jobs if they cannot pay the portion of
the insurance costs previously borne by their employer. As for employees of
small businesses exposed to the risk of closure and unemployment, it is very
likely that they will no longer have the means to take out a private insurance
policy on their own. Already, in early 2019, just over 9% of the population had
no health coverage. While this rate had dropped sharply since 2010 and the
“Obamacare” reform, the annual report
of the US Census Bureau published in November 2019 estimated that more than 29
million people had no coverage in 2019, a figure that has risen somewhat since
2017. The coverage rates also show strong regional disparities, which is due to
the demographic structure of the States.
Although part of the economic support plan
is devoted to food aid [8]
and some health expenses, the COVID-19 crisis will once again hit the most
vulnerable populations and widen inequalities that are already significant and being
deepened by the recent tax reforms of the Trump administration.
[1]
In terms of GDP, the share of States that have imposed lockdowns is in much the
same proportions.
[2]
Note that this survey does not show a significant difference between men and
women, even if women have a slightly fewer opportunities for teleworking: 28.4%
against 29.2% for men.
[4]
In Nevada, the exposed sectors represent 34.3% of jobs. This figure also
exceeds 30% in Hawaiï and is 23.7 % in Florida.
[5]
This is also the case of the District of Columbia due to the large presence of Federal
employees.
[6]
This amount is granted to households
receiving less than USD 75,000 (150,000 for a couple) per year. USD 500 is
awarded per child. The amount of the tax credit is regressive and falls to zero
for households with an income above USD 99,000.
[7]
See here
for our analysis of European and American strategies to deal with the crisis.
Between 2017 and 2018, Finland conducted an experiment with universal income that gave rise to significant media coverage. 2,000 unemployed people receiving the basic unemployment benefit (560 euros per month) received the same amount in the form of unconditional income, which could be combined with income from work for the duration of the experiment (2 years, not renewable). On 6 May 2020, the final report evaluating the experiment was published (here is a summary of the results). The evaluators concluded that the experimental universal income had moderate positive effects on employment and positive effects on economic security and mental health. According to the final report, on average individuals in the treatment group worked approximately 6 additional working days (they worked 78 days). They experienced significantly less mental stress, depression and loneliness, and their cognitive functioning was perceived as better. Life satisfaction was also significantly higher. The results of the experiment therefore seem to argue in favour of a universal income. But is it really possible to draw lessons from the experiment with a view to generalizing the system? In 2018, I wrote that experimenting with universal income was “impossible“. Does the Finnish experience contradict this claim? It turns out that it is indeed difficult to draw lessons.
The principle of a universal income, as it
is commonly defined, is to pay a sum of money to all members of a political
community, on an individual basis, without means-testing or any obligation to
work or take a job.
Such experiments generally concern a small
number of people (in Finland, 2,000 individuals): the universal aspect of the
measure is therefore lost, but a measure’s impact can differ depending on
whether it affects everyone or only some of the population. How are the individuals
chosen? Two options are favoured by practitioners: a totally random draw, which
favours the representativeness of the experimental sample, or a saturation site,
which consists of including in the experimental sample an entire community (for
example a single labour market area), which helps to capture externalities and
interactions (“do I stop working more easily when my neighbour stops or
when my spouse receives assistance?”). In Kenya, villages
are used as saturation sites. In the Finnish experiment, 2,000 long-term
unemployed people receiving end-of-entitlement benefits (equivalent in France
to ASS assistance) constituted the experimental group, with the control group
being made up of recipients of end-of-entitlement benefits who had not been randomly
selected. This poses two problems. First, the experimental group is not
representative of the Finnish population. The long-term unemployed make up only
a small part of the population. So we cannot really say how people with jobs would
have reacted (would they have reduced their working hours?). Second,
interaction effects are not taken into account: for example, consider a job taken
up by an unemployed person in the experimental group, who thus increases his or
her labour supply in the context of the experiment – might this job have been taken
up by a member of the control group?
The definition of universal income tells us
nothing about its level or what benefits it replaces. All options are on the
table. Programmes with a more liberal, free-market orientation offer a
relatively low universal income and replace most social benefits and sectoral
subsidies (notably in agriculture) or can even substitute for regulations on
the labour market (the abolition of the minimum wage is envisaged). In a more
social-democratic logic, universal income would replace only the social minimum
(France’s RSA income support benefit) and income support for the in-work poor
(in France, the Prime d’activité). The amount envisaged is often equal
to or slightly higher than the social minimum. Finally, in a degrowth logic, the
universal income could be lifted to at least the poverty line in order to
eradicate statistical poverty. The effects expected from the reform depend
greatly on the amount envisaged and the benefits it replaces. In the framework of
the Finnish experiment, the universal income was 560 euros, the amount of the
basic unemployment benefit received by the members of the experimental group. Simply
replacing this basic allowance meant that at first the income of the unemployed
in the experimental group remained unchanged. But the universal income could at
the same time be cumulated with job income. This means that returning to work could
lead to an additional financial gain of as much as 560 euros.
The experimentation thus increased the
financial gains from a return to work. This is not a result that one usually thinks
of in relation to establishing a universal income. One question often asked is,
“What
happens when you get 1,000 euros a month without working?” It turns
out that, for those on low incomes, the generalized roll-out of a universal
income could have ambiguous effects on the incentive to work: it increases
income without work but it also provides additional income for the working poor.
On the other hand, for those earning the highest incomes, the monetary gain
from increasing their income would be reduced.
The evaluation was complicated by the
introduction of activation measures during the second year of the experiment
(2018). Based on the “activation model” put in place, people on unemployment
benefits had to work a certain number of hours or undergo training, otherwise their
benefit was reduced by 5%. These measures affected the experimental groups
asymmetrically: two-thirds of the control group were affected, compared with only
half of the experimental group (Van
Parijs, 2020). Theoretically, the incentive to return to work was therefore
greater for the control group. Note that activation goes against the principles
of the universality and unconditionality of universal income.
Notwithstanding the activation measure, the
results of the Finnish experiment tell us that the hours worked are higher for
the experimental group than for the control group. The financial incentives to
work would therefore have worked! In fact, the evaluators stress the moderate degree
of the impact on employment. In the interim report, which covered the first
year (2017), the impact was not significant. In 2018, the impact was
significant, since the people in the experimental group worked an average of 78
days, or 6 days (8.3%) more than the control group. The impact is, however, not
very significant: with a 95% confidence interval, it is between 1.09 and 10.96
days (i.e. between 1.5% and 15%). Kari Hämäläinen concludes:
“All in all, the employment effects were small. This indicates that for
some persons who receive unemployment benefits from Kela [Finland’s agency
handling benefits for those at end of entitlement] the problems related to
finding employment are not related to bureaucracy or to financial incentives”.
On the other hand, the experiment tells us nothing about the effects of
possible disincentives for higher earners due to the financing of the measure:
by construction, an experimental universal income is not financed. More
seriously, gender analysis is virtually absent from the final report. All we know
is, from reading a table, that women in the experimental group worked 5.85
additional days compared to 6.19 for men, but there is no discussion of the
issue of gender equality. The issue of how choices are negotiated within a household
is also not posed. The impact on the lone parent group is not significant
“due to its small size”. In an Op-Ed
published by the New York Times, Antti Jauhiainen and Joona-Hermanni
Mäkinen criticize the sample size, which is five times smaller than initially
planned: the small size makes it difficult to draw any conclusions about subgroups.
The final report highlights the beneficial
effects on mental health and economic well-being. The impacts on people’s life satisfaction
and on stress and depression are very significant. However, two comments can be
made. First, we do not know what comes from the higher living standards of the
individuals in the treatment group and what comes from the mechanism of a universal
income (the certainty that people will have an income whatever happens). Given
the way the experimental income was actually designed (it functions like an
employment bonus), one can easily assume that it is the income effect that
takes precedence. Likewise, since the individuals in the experimental group are
in all cases better off financially, it is not surprising that their economic
well-being increases. Second, there may also be a reporting bias due to a Hawthorne Effect:
individuals in the experimental group know that they are part of an experiment
and that they were chosen so that they have an advantage over the control group.
This can lead them to be more optimistic in their statements.
In the end, the Finnish experiment offers
few lessons about the effects of the establishment of a global universal
income, i.e. one for all citizens. Only a small category of the population was
involved, and funding was not tested. Yet funding is half the mechanism;
Finnish trade unions are also opposed to a universal income because they fear
that the necessary tax increases will reduce earnings from working. In
addition, a family and gender approach has been completely ignored, whereas a universal
income has been denounced by feminists as being liable to discourage women from
taking up jobs (likening it to a mother’s wage). As with the RSA income supplement experiment
in France [article in French], the failure of the Finnish experiment is
explained in part by the contradictory objectives of the various scientific and
political actors. The evaluators hoped for a sample of 10,000 people including individuals
with different employment statuses. They were constrained by a combination of time,
money and a ruling political coalition that was no longer enthusiastic about
the idea of testing a universal income (“Why
Basic Income Failed in Finland”). The Prime Minister’s Centre Party
was in fact interested in the question of financial incentives for the
long-term unemployed, which is a long way from the idea of reconsidering the
central role of market labour or being able to say no to low-quality jobs, which
is often associated with universal income. This was certainly a limitation of
these costly experiments: subject to the inevitable supervision of politics,
they risk becoming showcases promoting the agenda of the government in power.
The second meeting of
the Eurogroup did the trick. The Ministers of Finance, after having once again laid
out their divisions on the issue of solidarity between euro area Member States on
Tuesday 7 April 2020, reached an agreement two days later on a fiscal support plan
that can be put in place fairly quickly. The health measures taken by the Member
States to limit the spread of the Covid-19 pandemic will enjoy better
short-term financing, which is good news. The additions to Europe’s tools for
dealing with the crisis will be on the order of 500 billion euros – this is
certainly not negligible, and note that this comes on top of the efforts
already put in place by governments – but this corresponds mainly to a new
accumulation of debt by the Member States. The net gain for each of them, as we
shall see, is actually quite marginal.
The Eurogroup will
propose the creation of a credit line (Pandemic Crisis Support) specifically
dedicated to the management of the Covid-19 crisis within the framework of the
European Stability Mechanism (ESM), without strict conditionality (meaning that
recourse to the credit line will not imply any control on the part of the EMS
over the future management of the Member State’s public finances). The creation
of the credit line was inspired by the proposal by Bénassy-Quéré et al. (2020), the advantages and disadvantages of which we presented to the Eurogroup meeting on
9 April 2020. The amount allocated to this credit line represents around 2% of
the GDP of each euro area Member State, or nearly 240 billion euros (in 2019
GDP).
The lending mechanism
proposed by the European Commission to supplement the partial unemployment
programmes of the Member States – it goes under the name of SURE – will clearly see the light of day and will be
endowed with 100 billion euros. For the record, the three main beneficiaries of
SURE cannot receive a combined total of more than 60 billion euros in loans.
Finally, the European
Investment Bank (EIB) will grant an additional 200 billion euros, mainly to
small and medium-sized enterprises in the EU Member States. In total, the euro area
countries will have 480 billion euros in additional financing capacity.
Table 1 below
presents a breakdown by country of the amounts in play. As part of the 240
billion euros of Pandemic Crisis Support, Germany will be able to benefit from
a borrowing capacity of nearly 70 billion euros, France nearly 50 billion
euros, and Italy and Spain 35 and 25 billion euros respectively. These amounts
correspond to 2% of the 2019 GDP of each country. At this point, there is no
indication of whether the Member States will draw on this capacity. The
advantage in doing so depends crucially on the difference between the interest
rate at which they can finance their health and economic expenses without using
the EMS and the interest rate on loans made by the EMS. The financing cost without
going through the EMS is the interest rate on the country’s public debt. The
cost of financing through Pandemic Crisis Support is the interest rate at which
this credit line is itself financed, that is to say, at the lowest rate on the
market, i.e. the German rate. So it is obvious that Germany has no interest in
using this credit line. Of the 240 billion euros allocated to Pandemic Crisis
Support, the 70 billion euros for Germany is thus useless. For countries other
than Germany, the use of Pandemic Crisis Support depends on the difference between
their interest rate and Germany’s rate, the infamous spread. If the spread is
positive, using the EMS effectively reduces the cost of borrowing. But as shown
in Table 1, the gain enabled by Pandemic Crisis Support is rather low. For
Greece, whose spread vis-à-vis Germany is the highest in the euro zone, the
gain would come to around 0.04% of GDP in 2019, i.e. a 215 basis point spread
multiplied by the amount allocated to Greece for Pandemic Crisis Support (3.8
billion euros, which corresponds to 2% of its GDP of 2019), all relative to its
2019 GDP. For Italy, the gain is on the same order: 0.04% of its GDP. Expressed
in euros, Italy stands to gain 700 million euros. For France, whose spread
vis-à-vis Germany is much lower than that of Italy, the gain could be 200
million euros, or 0.01% of its GDP in 2019.
Assuming that the amounts allocated by the EIB are prorated to the country’s size (measured by its GDP in 2019), and that Spain, Italy and France benefit from 20 billion euros each under SURE, the total interest rate savings would reach, respectively, 680 million, 1.5 billion and 430 million euros (0.05%, 0.08% and 0.02% of GDP). At a time when it seems to be raining billions, these are not big savings. Unless you think of it as a metaphor. Like rain before it falls, the billions of euros are not really euros before they fall.
