3.03.2021 57

Covid, productivity and long-term growth: further secular stagnation or positive reversal?

Published March 3, 2021

5 to 10 minutes

5 to 10 minutes


While the spectacular Covid shock may easily lead to exaggerate extrapolations, it cannot be ruled out that it will matter for long-term productivity and growth. However, the effects could work both ways. Reasons why the current crisis could further worsen “secular stagnation” are many, yet there are also a channels through which it could work positively.

Since the 1970s, productivity gains in the developed economies have gradually slowed (from around 2% per year to less than 1% in the 2010s), despite intermediate cyclical rebounds (at the end of the 1990s in particular).
Several reasons have been put forward to explain this slowdown, some linked to supply (de-industrialisation, the low contribution to productivity from information technologies), others to demand (ageing, loss of confidence following crises, over indebtedness, inequality) and others still to measurement issues (difficulty in factoring in the increasingly rapid appearance of new products and services).
It cannot be taken for granted, at this point, that the current crisis will have a long-lasting effect, positive or negative, on productivity growth (as well as on the other components of the broader economic “secular stagnation”). The spectacular 2020 shock may have created an environment conducive to excessive conclusions about its capacity to alter well-entrenched trends that stretch back several decades and are underpinned by many factors.
Because something is uncertain, however, does not mean it is impossible, and some aspects of the Covid crisis do seem enough to warrant reflection at the very least on their potential longterm effects on economic growth. Lasting changes could affect each component of a traditional production function: Total Factor Productivity (TFP: productivity of both labour and capital), capital inputs and labour inputs. Some of these changes maybe negative and some positive.

From the perspective of TFP, some negative consequences of the crisis may indeed affect long-term productivity growth, while others could be limited to “level” (one-off, even though permanent) effects:

  • The possibility that the crisis will lead to social tensions or political changes that reduce the likelihood that governments can or wish to implement “pro-supply” policies. Such policies, which generally involve reducing the protection of incumbent players against new competitors on the labour, goods and services markets, are widely considered to generate lasting positive incentives for productivity.
  • The potential acceleration of deglobalisation, making companies less exposed to international competition, and therefore with less incentive to permanently improve their processes o maintain competitiveness, and less opportunities for massive economies of scale.
  • The “zombification” process associated with prolonged public support for nonviable and uncompetitive companies, which would lead to protracted suboptimal allocation of capital and labour resources, and possibly also a deterioration in skills. Note, however, that while often mentioned, this process would need to remain several years to have more than cyclical and level effects (i.e. slowing down the recovery after mitigating the recessionary shock, and causing one-off skill attrition) and really alter long-term productivity growth. If maintained over the long-term “zombification” could then be very much part of the first channel (lack of pro-supply policies).
  • The possibility that the disruptions to the educational system during the crisis will have a negative impact on professional skills when current students (or school dropouts) enter the labour force. However, this is another factor that may concern more the level of productivity a few years after the crisis than its long-term growth (indeed, while it is well documented that a more educated population is more productive, it is more debatable whether it can grow its productivity faster over time).



Pro-supply economic reforms may be difficult to implement in the post-crisis social and political environment.

Education disruptions during the crisis may have lasting consequences.

Teleworking could have significant implications on housing and mobility bottlenecks.

From the point of view of capital inputs, at least two factors could have a negative impact:

  • The high levels of debt, together with durably shaken confidence could hinder risk-taking and investment and therefore the long-term pace at which new capital is built.
  • An increase in bankruptcies as we exit the crisis would add a negative level shock through the premature attrition of productive resources that may not find other takers.

Concerning labour inputs, at least three negative drivers may also be identified:

  • Less immigration linked to the fact that professions employing a high proportion of immigrants have been significantly impacted by the crisis yet also, on a more lasting basis, by the possibility that social or political tensions following the crisis could lead to anti-immigration policies.
  • A level shock could be added by workers permanently shifting out of the labour force after the prolonged Covid-related inactivity period.
  • The crisis could also have negative consequences for the birth rate, with a long-term impact on the active population.

Finally, in addition to supply factors in the production function, it is possible that growth will also be durably affected by structurally low demand. This could stem (as with the reduction in investment) from more prudent behaviour trends (higher saving rates) due to low confidence or high debt levels. It could also stem from an increase in social inequality (greater concentration of revenue and wealth among populations with a high savings rate).

That said, there are other effects of the crisis that could favour long-term growth, notably via an increase in the TFP.

In this case, we can also identify several factor categories:

  • The opportunity provided by the crisis of testing and developing multiple “disruptive” new technologies and forms of organisation. Massive use of online work, education and shopping, as well as medical innovation, are the most obvious illustrations. In reference to innovation cycle theory, it is notably possible that the crisis speeds up the widespread “diffusion phase” of information technologies which had previously been in “discovery phase” and, therefore, had not yet generated the bulk of their potential contributions to productivity.
  • Some extended potential implications of home working (in addition to the gains or loss in productivity of the employee concerned) on businesses (lower costs, access to a much broader pool of skills, new relocation and offshoring possibilities) and on territorial organisation (potential opportunities to circumvent transport congestion and housing shortages that impair work and skill mobility).
  • The possibility that major public stimulus plans will, in addition to their demand-supporting role, effectively target innovation (an ambition clearly stated in the case of the European NGEU plan, for instance).
  • A potential “natural selection” effect through the elimination of weaker companies (presumably less productive) once the public protective measures are discontinued. This process could, however, be less relevant than during past crises, since Covid is an exogenous factor that has impacted companies primarily, even those that are well managed, depending on their sector of activity.

As such, in addition to bringing a number of negative one-off shocks to TFP, capital and labour, the Covid crisis could carry very contradictory forces for long-term productivity and growth. It is probably too early to drawconclusions as to their strengths and even more so their net effect, particularly since the Covid crisis is not yet over (if it were to become protracted, causing more damage yet also more incentives for adaptation, its long-term growth effects would probably only become more ambiguous).

It seems nevertheless fair to assume that the policies implemented by governments in the immediate aftermath of the crisis will play a significant role in this tug-of-war between positive and negative long-term consequences. In light of the points set out above, the main challenges for public authorities in extracting as much positives as possible are likely to be 1/ keeping some “pro-supply” reform momentum alive; 2/ ensuring that substantial support for demand can be maintained without crystallising rigidities; 3/ ensuring that stimulus plans are at least partially directed towards harnessing productive “disruption” (against a backdrop where there will probably be strong political or lobbying pressure to use them to other ends).

Our online help service is available to answer your question

My personal information

If you have a question about our company or one of our products, please complete the form to get in touch. Please do not mention your account numbers or critical data in this form.


(*) Required fields
All our job offers (Permanent and temporary position, Internship, Apprenticeship or VIE) are available on our dedicated website:

Register and apply directly online.

Amundi on Twitter