Editors' note: This column first appeared as a chapter in the eBook Publishing and Measuring Success in Economics, edited by Sebastian Galiani and Ugo Panizza.
At the end of March 2020, CEPR issued a call for papers for Covid Economics, Vetted and Real-Time Papers. Within a couple of days, submissions started to pour in. At the time of writing (26 June 2020), 485 submissions had been received, 219 papers had been accepted, and 32 issues had been posted online, freely accessible.
The immediate reason for launching this new form of collecting articles was an outpouring of submissions to VoxEU, CEPR’s portal for short, policy-relevant contributions. From early March, it became clear that economists around the world, like everyone else, were mesmerised by the pandemic and trying to make sense of the unfolding events. A cartoon at the time showed a man saying: “All my economist friends used to be climate change experts, now they all are epidemic experts”. Within two weeks, CEPR published two eBooks. Opinion pieces, informed by standard economic principles, were soon followed by elaborate original research, both theoretical and empirical. It was clear that many economists, coming from all areas of the discipline, were working hard. Hence the idea of bringing together their work in a freely accessible website.1
The inspiration came from physics and the medical sciences, where there exists an old tradition of pre-prints – working papers that are lightly refereed and posted very fast. Likewise, Covid Economics vets submissions on an accept/reject basis. About 30 researchers from all subfields were invited to join the Editorial Board with the explicit requirement that they review submissions within 48 hours.2 The Editorial Board includes established researchers who can credibly decide ‘up’ or ‘down’ in a short period.3
The accepted papers are then collected in ‘issues’ with a frequency and size endogenous to the flow of accepted papers. Table 1 indicates the time lags between submission, decision and posting.
Table 1 Time lags (in days)
Note: Yes, the distribution is truncated and this is not taken into account.
Issue size and frequency were soon boosted to three issues per week with eight papers in each. Once lockdowns came to an end, the number of submissions started to decline, as shown in Figure 1. Like everyone else, economists have started to think about something other than Covid.
Figure 1 Daily submissions, 26 March to 26 June 2020
Note: includes weekend days.
Submissions have been received by authors based in 47 countries. Table 2, which lists the top 20 countries, reveals a few observations. First, there were only two submissions from China (both rejected) and fewer from France than could be expected. One can only speculate as to why China-based scholars are not active. Second, most of the authors (86%) work in academia, with most of the rest at central banks. Third, only 17% of the authors so far are female, a point already made by Amano-Patiño and co-authors in this eBook. Several papers featured in Covid Economics report that women with children find it difficult to work from home, an observation that may well apply to researchers during the lockdown period.4 Fourth, a large number of excellent papers are submitted by junior researchers and some from developing countries. As intended, through Covid Economics, CEPR is serving the whole profession.
Table 2 Home base of authors
Note: Data for period 26 March to 26 June. Many papers have multiple authors. International refers to authors from international institutions (IMF, World Bank, etc.).
Fifth, while a number of leading economists have spearheaded the research efforts on the pandemic, many have chosen to issue their work in existing working paper series (NBER, CEPR, their own institutions or research centres) rather than through Covid Economics, even though they have sometimes been invited to do so and we never ask for exclusivity.
A new concept in economics
As other contributions in this eBook indicate, the publication time lag in most professional reviews is frustratingly long. A good case can be made that academic research requires care and serious refereeing, and that the ‘revise and resubmit’ process greatly improves quality. Yet, when stunning events occur requiring rapid policy reactions, this process becomes wholly ill-suited. It can be argued that even in normal times, the long lags are too long.
One solution has been the widespread circulation on the web of discussion papers issued by well-known networks such as CEPR and NBER, as well as international institutions and central banks. However, in the case of most of these series, only fellows or insiders can publish and they may only be available behind a paywall. In addition, there is a big difference between publishing a CEPR or NBER working paper and one in a departmental series. NBER and CEPR are closed shops for a very small group of researchers. This means that there already is a quality filter at the fellow level, applied at the time of appointment. Furthermore, NBER, CEPR and some institutions also vet the discussion papers.
A key innovation of Covid Economics is that it has opened up CEPR for quality publishing beyond its own network while ensuring a high standard of quality. In doing so, Covid Economics has inherited some of the reputational capital of the CEPR network. Very few departments or institutions could have done this. CEPR is covering all the related costs, truthful to its not-for-profit mandate (which is possible in part because it is selling its normal working papers).
As they explicitly own the copyright, authors may also issue their works in other working paper series, and many do so. (Some also submit their papers to medical pre-print series.) From CEPR’s standpoint, Covid Economics is very explicitly a public service effort – to provide researchers with real-time information by collecting quality contributions.
It was probably inevitable that an alternative approach would emerge one day. The Covid crisis provided the first impetus. The creation of Covid Economics may be seen as a challenge to the long-established tradition of journals as well as the dominance of working paper series. This is not the intention. From the start, it was decided that its contents would be akin to discussion papers, in the expectation that many featured papers would later appear in improved versions in the regular journals. Indeed, the vetting process of Covid Economics implies that the papers are accepted as submitted. In some cases, the reviewer would point out weaknesses (argumentation, wording, language quality) that can be promptly taken care of. Some reviews are more extensive, in which case the authors are told that it can help them to improve their papers for future submissions elsewhere. In extremely rare cases (fewer than 1%), the authors are asked to remedy non-fatal weaknesses, provided this can be done in a few days.
In order to facilitate future journal publication, we contacted a large number of the top journals to ask them whether they would accept submissions of papers previously featured in Covid Economics. The reaction has been overwhelming: all of the contacted journals have agreed (some of them have indicated that they would require significantly improved versions) and a few others volunteered. The list of the corresponding 29 journals is indicated in every issue.
During these exchanges with the journals’ editors, a question has come up: What is a journal and why is Covid Economics not a journal? Journal characteristics may include its typesetting. The first two issues were indeed uniformly typeset, but this was abandoned when we realised that the flow of submissions was such that it was becoming impossible to continue and still be ‘real time’ as intended. Another question is whether the authors themselves consider it as a journal and mention it in their CVs. Of course, this is not something that we can control but, as result of these discussions, the acceptance mail now makes it clear that we consider the contributions as a working paper.5 CEPR’s position has always been that a journal is characterised by its extensive refereeing process, while Covid Economics actively bans ‘revise and resubmit’.
Editing Covid Economics is a very special undertaking, quite unlike editing a standard journal. The volume of submissions is one aspect, as is the reviewing process which calls for ‘real-time’ decision making.
On an average day, including during weekends, I have been receiving about six submissions and six review reports, with a wide variation as Figure 1 makes clear. The endogenous and real-time nature of the operation implies reacting on the same day. Obviously, I cannot read in detail each and every paper, but I have to do so carefully enough to determine whether to send it for review or reject it at once, and whom to send it to.
Submitted papers may revisit issues already treated or rely on methodologies previously used. Over time this has happened more and more, and the reviewers may not be aware of all relevant publications while, gradually, I have lost the ability to precisely remember each and every paper. This requires sifting through the list of accepted papers, looking at some again and passing the information to the reviewer.
The reviewers are asked to make decisions, but I need to check that these decisions are coherent from one reviewer to the next. The reviews are meant to be short, and some are, and they must be decisive. Over time, increasingly often, the reviewers state that they are ‘sitting on the fence’. This requires clarifying with the reviewer his/her view and agreeing on a decision, without taking too much of their time given the tight deadlines under which they are required to operate. Then, of course, comes the occasional reaction of the authors of rejected papers, which I usually do not have the time, nor inclination, to respond to.
A 'new' field
Like everyone else, economists have suddenly seen their lives deeply disturbed by Covid-19 and have been mesmerised by the idea that governments could close down large swathes of economic activity. From all corners of the profession, many dropped what they were doing and started to use their tools to try and understand ongoing events and to explore the policy responses.
Prior to Covid-19, the field of pandemic economics was quite small. There was some research on the economics of epidemics, with a dedicated journal (Epidemics and Economics), but epidemics was mainly a field within the medical and biological sciences. The sudden interest of economists in pandemics has boosted research, creating an almost new field, largely ignoring previous work. Progress has been swift, both from a theoretical and an empirical angle. It is impossible to describe all these advances. This section merely attempts to present the main themes and results.
During the first few weeks, a significant share of submissions started from the standard epidemiological (SIR) model designed to evaluate the speed at which contagion spreads and kills. While epidemiologists have enriched their models over decades by tracking down contagion with more granularity, there was a big hole: lockdowns and other health measures have massive economic implications, some of which in turn affect the epidemic. Papers started to add an economic block to SIR models or to embed the SIR model into economic models. Some have also exposed some limits of the SIR models, for instance by showing the powerful effects of ‘superspreaders’.
From there, the natural next step was to look at a possible trade-off between health measures and their economic impact, dubbed ‘health versus wealth’. Various approaches were used to evaluate this trade-off, looking at policy optimality and evaluating the policies in place. Progressively, attention moved to the evaluation of the health and economic impacts of various policies, from lockdowns, to face masks, to testing and tracing, and to working at home.
These feedback loops between contagion and economic conditions represent a major step in understanding the spread of Covid-19 in the presence of health policy decisions. Most studies find that policy measures are each useful to contain the virus spread, largely as complements to each other. Given the economic costs of lockdowns, many papers seek to describe what the optimal mix would be. To that effect, some use the value of statistical life concept, but others studiously avoid its use as they look at ways of minimising the economic costs for a given level of contagion or casualties. Many papers use various large databases that provide real-time measures of people’s mobility, which is found to decline quickly either as governments decide measures or as people become concerned by contagion.
It soon emerged that lockdowns, while effective, may not be cost-efficient when they are all-encompassing. It is likely that it would be better to target special segments of the population – those most at risk (older people and those with specific pre-existing conditions) and those who have tested positive and their contacts – or to impose the measures in particular geographic areas rather than in entire countries.
Working from home has become a lively topic of its own. Starting from earlier contributions, research has quickly moved on to determining who can work at home, which depends on jobs, education levels, geographical location, use of public transport and gender. It has naturally led to the documentation of the deleterious effect of the epidemic and health policies on inequality.
A virus, it might seem, affects all people equally, but a large number of papers have shown that Covid-19 has sharply deepened pre-existing inequalities, for many reasons. Within countries, social distancing has been a sort of luxury for poorer hand-to-mouth workers. Lockdowns have exacerbated differences in housing amenities and transport, not to mention access to health facilities and mental health effects. Across countries, the less- developed ones face daunting challenges given their large informal sectors, limited health provision, meagre social programmes, restricted ability to borrow publicly and privately, and more.
Big data have allowed researchers to analyse the effects of lockdowns on people’s mobility. They have started to use highly granular information on individuals’ movements to examine the mobility response to lockdown orders, changing shopping habits, consumption patterns, price formation, as well as the reactions of asset markets to news. They have related these observations to people’s occupations, incomes, location and even political preferences, revealing rising inequalities. There is also some evidence that lockdown orders may add little to spontaneous isolation when people become aware of the contagion spread. Surprisingly, perhaps, big data have been available for emerging countries, not just developed countries, which has made it possible to engage in comparative analysis.
An important issue concerns how the characteristics of societies affect their ability to deal with the pandemic. The amount of social capital is found to a play an important role, as do political preferences, religiosity, culture and existing trust and norms. A few papers also tackle the question of whether democracies are more or less effective than dictatorial regimes. At this stage there is no consensus, in part because of data limitations.
Many papers have looked in some detail at the economic impact of the pandemic and, in some cases, attempt to make predictions using a wide array of techniques. Real-time measures of consumption from large databases have quickly shown the depth of the contraction, although not all sectors or sale channels have been affected in the same way; some have been large beneficiaries. Policy evaluations have documented the role of cash transfers, unemployment benefits, bailouts of firms, and so on in mitigating the adverse economic effects of lockdowns.
Financial markets are also fertile ground for analysis since the movements have been large and real-time data are easily available. As was likely, the stock and bond markets have responded to daily news, especially about casualties, and to policy measures, with some evidence that central bank interventions have been effective in preventing a financial crisis. A couple of papers even indicate that the stocks of firms with strong environment and social governance (ESG) have fared better than the others. Other papers have studied early events of illiquidity and examined policy responses as well as the impact of existing regulations.
Some papers explore how individual people react, using survey data, laboratory experiments or even data on calls to emergency helplines. They show that mental health is seriously disturbed, often with stronger effects on women.
Contrary to what could be expected, economic historians have made relatively few contributions. For time immemorial, global pandemics have repeatedly brought havoc and major changes. Economic historians have painstakingly recovered data to analyse these events long before Covid-19, but only a limited number have submitted articles that could inform us on the lessons to be learned from previous pandemics.
The agenda is still expanding. The removal of lockdowns is proving to be hectic, leading to new measures. The long-run impact of Covid-19 remains to be analysed. The organisation of healthcare, work and family life have been challenged and could be different from what we have known once the danger of contagion has gone.
Impact on media and policymakers
Many researchers shifted their attention to the pandemic because they sensed an urgency to properly understand an historical event. The creation of Covid Economics was related to the same perception that new knowledge had to be promptly developed and communicated to policymakers, who faced sharp choices.
Assessing how academic research shapes policymaking is always very difficult. Except in very rare cases, results trickle down slowly via a variety of channels. For many weeks, policymakers and the media have focused almost exclusively on Covid-19. A huge amount of information has circulated and been used to shape policy decisions. Early on, most of the relevant information was provided by epidemiologists and virologists, who understandably argued in favour of lockdowns since “it isn’t the virus that circulates but people”, to quote Swiss virologist Didier Pittet (Le Temps, 19 March 2020). It did not take long, however, to realise that the economic (and social) costs of lockdowns are vertiginous.
There may be a trade-off between health and wealth, and economists are uniquely trained to detect and deal with trade-offs. They are also well versed in distinguishing the long from the short term.
As noted above, much of early research in Covid Economics was dedicated to the theoretical analysis of potential trade-offs; as data were collected, research shifted to empirical evaluations. The media soon picked up on this theme. Over the weeks, further results from featured papers reached the media as well: the adverse effects of lockdowns on income and gender equity, more subtle approaches to social distancing, the role of politics in shaping policies and individual behaviour, analyses from big data on individual mobility, consumption or compliance with distancing policies, and more. Somehow, research was listened to.
CEPR has deployed all its resources to disseminate the wealth of results featured in Covid Economics. Each new issue is circulated to the Centre’s mailing lists and regular summaries of selected contributions are published to create awareness of their results. A new website dedicated to Covid-19 has been created and is used to facilitate access to the issues.
Overwhelmed by the task of managing a flow of submissions and publishing three issues per week, and sharply focused on research, Covid Economics did not have the human resources to promote papers with the media and policymakers. CEPR’s popular policy portal, VoxEU.org, routinely invites selected authors to write up widely readable presentations of papers featured in Covid Economics. At the height of the crisis – if the height is indeed a thing of the past – VoxEU published a record number of columns daily and its readership has risen by about 250%. In the second quarter of 2020, VoxEU had over 3.5 million page views (up 163% on the same period last year). In parallel, CEPR has published several eBooks,6 organised webinars and posted video interviews of economists, drawing heavily on papers featured in Covid Economics. Clearly, no economic journal can mobilise such an array of outreach tools. In effect, Covid Economics has become the centre point of CEPR activities since the end of March, even though the efforts are hampered by the sheer volume of papers coming through.
The future of Covid Economics
As previously noted, the launch of Covid Economics was motivated by the perceived urgent need to develop relevant economic research at a historical juncture. Little thought was devoted to the duration of this undertaking; it would endogenously respond to supply while ensuring that there is demand. Figure 1 shows that the number of submissions reached a high plateau in April and May but even towards the end of June remains quite elevated. The content of papers has gradually shifted as the new field is becoming more diverse and, predictably, more detailed. It is also following events, including the lifting of lockdowns and policy measures. The current conventional view is that the epidemic will last until a vaccine is found and administered universally, which could take another year or more. It is likely that the research effort will remain vigorous as long as the epidemic lasts. The end of the epidemic will also provide new data that will need to be analysed in real time.
Beyond that, it is probable that interest will wane. Yet, many researchers will have developed a specialisation and the prospect of other catastrophic epidemics could linger. Hopefully, however, urgency will come to an end. Does this imply that the need for real- time and freely accessible dissemination of research will also come to an end?
As already mentioned, the experience with Covid Economics has brought to the surface well-known complaints about the long lags between submission and publication. Standard publications may speed up the evaluation process (many have already done so for articles dealing with the pandemic). Indeed, one lesson from the Covid Economics experiment is that it is possible to ask referees to respond quickly. Of course, the 48-hour requirement is extreme, directly tied to the emergency of the situation, but short lags are possible without compromising the quality of the process.
Thus, beyond its contribution to a faster understanding of the pandemic, the Covid Economics experiment should help the profession think about how research is published. It has innovated in several dimensions:
- A new process, vetting, which can usefully complement traditional refereeing for quick dissemination of results. It ensures quality thanks to an Editorial Board composed of established researchers.
- A new format, somewhere between a standard publication and working papers. This is achieved by collecting papers in issues, without typesetting. It is cheaper than journals and costlier than working papers.
- New forms of dissemination: the link with VoxEU provides visibility beyond academic researchers, summaries of issues through CEPR, live presentations at the CEPR–Graduate Institute webinar.
- A new channel for publication, more inclusive than working papers and more attractive to younger researchers.
- From the point of view of the sponsor (CEPR) and the editor, Covid Economics is a public good, not a business model.
Authors’ note: I am grateful for real-time comments and suggestions from Giancarlo Corsetti, Barry Eichengreen, Antonio Fatas, Francesco Giavazzi, Warwick McKibbin, Ugo Panizza, Richard Portes, Tessa Ogden, Barbara Petrongolo and Beatrice Weder di Mauro. I am indebted to the members of the Editorial Board of Covid Economics, whose dedication and clear-mindedness have made this experiment not just possible but also highly successful.
1 The idea was conceived by CEPR President Beatrice Weder di Mauro and VoxEU Editor-in-Chief Richard Baldwin.
2 As the number of submissions rose to more than six per day on average, and twice as many on some days, the Editorial Board was enlarged to its current size of 51 members.
3 It may be noted that, as a result, Covid Economics’ ‘vetting’ implies high quality judgement, even in comparison with established journals that occasionally rely on fairly junior referees.
4 Women members of the Editorial Board of Covid Economics account for about one-third of the total.
5 The exact statement is: “Please note that CEPR treats papers in Covid Economics as it does with its Discussion Papers: authors retain the copyright and can submit the paper to journals that accept submissions of papers previously featured in Covid Economics. A non-exhaustive list of such journals appears in the prelims of each issue of the publication from issue 16 onwards. If your paper subsequently appears in a journal, this can be mentioned retroactively if you let us know.”
6 In fact, two ebooks, edited by Richard Baldwin and Beatrice Weder di Mauro, were the first economic publications on the pandemic (Economics in the Time of Covid on 11 March and Mitigating the COVID Economic Crisis: Act Fast and Do Whatever It Takes on 18 March).