Claire Leroy – PhD student in economics at CREST, École polytechnique, IP Paris

In September 2025, she will be a post-doctoral fellow at the Institute for Fiscal Studies (IFS) in London for one year. She will then take up, in September 2026, a position as assistant professor at LMU University of Munich.

Her research interests are in public economics and applied microeconomics.

She studies the impact of public policies on the economy and ways to improve their design, using empirical and theoretical approaches. She focuses on the role of imperfect information and other frictions in how citizens perceive and react to policies.

Her thesis, Raising Take-up of Welfare Programs: Evidence from a Large French Reform, was supervised by the economists Pierre Boyer and Antoine Bozio.

1. What have been the key stages of your research journey?

I began my thesis at CREST in September 2020, in the middle of the Covid crisis. As a result, my first year was a bit special: most activities, like the weekly seminars, were held online. There was little contact with other members of CREST because access to the department was still quite limited. From the end of the first year, things started to return to normal. And from that point on, my thesis was anything but solitary. I really began to build ties with other PhD students and researchers. Looking back, I think that this first “Covid” year made me realize how essential exchanges are—whether informal discussions between PhD students or seminar presentations—in research work.

A key moment of my thesis was my five-month research stay at the University of Berkeley. There, I was able to interact with Emmanuel Saez, the professor who had invited me, as well as with other researchers and PhD students. Given the level of this university, the atmosphere really pushes you to be ambitious in your research projects. This stay allowed me to meet many people, in particular other visiting PhD students, and to build a real network. I still come across people I met there at conferences.

But the experience that perhaps marked me the most was the “job market” period—that somewhat mythical moment everyone talks about during the PhD. You often hear that it’s super stressful, complicated… For my part, I ultimately found it much easier to navigate than expected. I think the mutual help between PhD students played a big role, and the atmosphere at CREST as well: it really is an environment where candidates are well supported in this stage.

2. Did you benefit from a working environment favorable to the advancement of your research—seminars, working groups, collective research activities?

CREST offers very good working conditions for PhD students. On the material side, I was able to obtain funding to do a research stay abroad, to attend conferences, and to obtain an additional year of doctoral contract.
But what I found just as important is the general atmosphere of the department. It is a very friendly environment, with little hierarchy compared to other departments. Within the economics department, everyone knows each other. There are not really barriers between PhD students and professors—it is quite normal to have lunch together, for example. There is also a real spirit of mutual help among PhD students, without competition. That makes the PhD, which can be a somewhat difficult period, much more livable. For several years now, PhD students have been organizing “PhD Breakfasts,” where useful information is shared with newcomers (about teaching, the job market, research stays abroad, etc.). And there are plenty of other initiatives in the same spirit, which are quite unique to CREST.

What makes it work, I think, is that people often come on site, mingle, and there is a real willingness among PhD students to get involved in organizing these moments. From cohort to cohort, PhD students pass on tips, advice, and often encouragement. I think this atmosphere and this collective involvement is a very precious intangible asset… but also a bit fragile. Preserving it really requires everyone’s involvement.

3. Did the PhD confirm or change your initial professional plan?

Before starting my thesis, I worked for three years at the Institut des Politiques Publiques (IPP) as a junior economist. I had already had in mind the idea of doing a thesis since my research master’s, but at the time, I had not found supervisors with whom it would have been possible.

When I began the thesis, I told myself that if an opportunity for an academic career arose, I would be delighted to seize it, while being aware that positions are rare and that it is a demanding path. I was therefore also open to the idea of continuing in a more applied direction, similar to what I had known at IPP. I approached the thesis as a kind of test: I wanted to check whether academic research really corresponded to what I was looking for.

At the end of the doctorate, I finally chose to apply for professor positions, because this test proved conclusive. Even though research obviously involves challenges, I felt in my place throughout the thesis, and this experience confirmed my desire to commit to a career as a teacher-researcher.

4. What advice would you give to a PhD student at the beginning of the journey?

I think one can give a lot of advice on very practical aspects of the thesis, but one of the most important undoubtedly concerns PhD life in general. As we know, it is a period that can be demanding, and many PhD students encounter at one time or another difficulties that can weigh on mental well-being. There is no magic recipe, but just as we invest time and energy in our research, my advice is to realize that living one’s thesis well requires personal work that also takes time and effort: learning to identify moments of blockage, knowing how to seek support when necessary, reflecting on one’s own ways of functioning. I think this work, which we don’t talk about much, is fully part of the doctoral journey.

For me, the PhD is a test period that allows you to see whether you can do research without the ups and downs of the profession—uncertainty, self-doubt about one’s abilities, stress—taking up too much space. And if it becomes too difficult, one must also know that ending one’s thesis is a completely legitimate decision, which we should normalize more. That said, I think it is important to deconstruct the very widespread idea that a “normal” PhD student must necessarily be suffering. That is not true. A thesis can also go very well, with of course a few harder moments, but overall, it can be an extremely rich experience.

5. How would you define an “in-work benefit” and what is its main objective within social policies? In what way is this type of measure important in combating precarity and improving the well-being of beneficiaries?

Two of my thesis articles deal with in-work benefits, a particular type of social assistance that appeared in the 1970s, first in the United States with the establishment of the Earned Income Tax Credit (EITC), then adopted in many countries from the 1990s onwards. In France, these measures have taken several forms: the Prime pour l’emploi (PPE) in 2001, the RSA “activité” in 2009, then the prime d’activité, which exists today and replaces these two other measures. This type of aid, which sometimes takes the form of tax credits, has a dual objective: reducing poverty while encouraging beneficiaries to remain in or return to employment.

Unlike traditional social assistance, such as the RSA in France, which targets people without any resources—and therefore without professional activity—in-work benefits are aimed at the working poor, that is, people in employment but with low incomes. By their design, these measures seek to make work financially more attractive, ensuring that “work pays.” They thus aim to limit what are called “poverty traps,” often associated with traditional benefits—situations, in theory, in which individuals are discouraged from returning to work because benefits reduce the net gain from employment.

That said, the incentive effects of these benefits depend on the economic context. They are relevant when unemployment or inactivity is explained by a lack of labor supply on the part of individuals. Conversely, if difficulties in accessing employment come mainly from a lack of demand on the part of firms, these considerations become secondary. More generally, we observe that the two types of aid—traditional aid and in-work benefits—often coexist within the same socio-fiscal system. Moreover, several works in theoretical economics show that optimal social policy generally rests on a mix between these two approaches.

6. In France, what were the effects of the 2019 reform of the prime d’activité? What conclusions can be drawn from your analysis regarding the concrete impact of this reform?

The prime d’activité, which is part of the family of in-work benefits, is a social benefit paid under means testing to households in which at least one member is in employment. It typically targets people paid at the level of the minimum wage or those with paths marked by alternations between employment and inactivity, with unstable incomes.

At the end of 2018, in the context of the Yellow Vests movement, the government announced a major reform of the prime d’activité. The reform consisted of increasing both the amount of the benefit and the number of eligible households. In the second chapter of my thesis, I focus on a very striking phenomenon: the explosion in the number of applications from January 2019, with nearly one million additional applicants. My objective is to understand what explains such an influx. The results show that this increase cannot be explained solely by the mechanical effects of the reform, that is, the broadening of eligibility criteria.

Alongside this mechanical effect, a take-up phenomenon is observed: many families who were already eligible for the prime d’activité before the reform, but who did not claim it, finally applied. This is what we call a reduction in non-take-up. According to a 2017 estimate by the DREES, about 27% of eligible households did not receive the benefit. My study shows that this rate was halved following the reform. What is interesting is that this very significant effect was unexpected, since it was not the stated aim of the reform. Faced with the increase in public expenditure generated by these new beneficiaries, the State had to revise upwards the credits allocated to the prime d’activité.

7. In your view, what are the main obstacles to the effective take-up of social benefits, despite their proven effectiveness? Are there structural or behavioral factors that explain this phenomenon of non-take-up?

Research, in economics as in other disciplines such as sociology, has highlighted a large number of factors that can explain situations of non-take-up of social benefits. Three main categories of obstacles are generally distinguished: transaction costs (linked to administrative procedures, the complexity of forms, which are costly in terms of time but also psychologically for beneficiaries), stigma (the feeling of shame towards oneself or others associated with receiving aid), and information problems (lack of awareness of the scheme’s existence, poor understanding of eligibility criteria). These factors are often interpreted as belonging to the individual sphere: it would then be the action—or inaction—of potential beneficiaries that explains non-take-up. Some situations of non-take-up can also be attributed to decisions by the administration in charge of the scheme: for example, a wrongful refusal to grant aid, whether by error or deliberately.

These obstacles to take-up vary greatly depending on contexts and schemes. In the case of the prime d’activité in France, sociological work shows that some traditionally important barriers seem to play a more limited role. Stigma, for example, appears relatively weak. This benefit is aimed at workers and is generally seen as legitimate compensation in the face of wages deemed insufficient. Moreover, the administrative costs associated with applying for the prime d’activité are relatively low. The procedure is entirely dematerialized and does not require supporting documents. On the other hand, other, more indirect forms of costs may play a role. Beneficiaries must, for example, log back in online every quarter to declare their resources, a step that some forget or consider constraining. We also observe a high frequency of overpayments (amounts unduly received that beneficiaries must repay) and arrears (sums due but paid late). These situations, often sources of stress or misunderstanding, can discourage some potential beneficiaries from applying for the benefit, for fear of finding themselves in complicated administrative procedures.

8. Your study shows that the increase in the amount of the benefit had a limited effect on take-up, unlike an “information shock.” How do you explain this gap? By what means did you measure the impact of media coverage on the behavior of potential beneficiaries?

In the case of the prime d’activité, my work seeks to understand why some families began to apply for the benefit after the 2019 reform—and thus, by extension, what initially explained their non-take-up.
A first hypothesis is that of transaction costs. If non-take-up stemmed from a rational cost-benefit calculation—in other words, if families judged that the administrative procedures were not worth it given the amount of the benefit—then the increase in the amount paid should have encouraged some of them to take the step. But my empirical analysis allows me to reject this hypothesis.

What I show, on the contrary, is that the increase in take-up is explained mainly by an informational shock. This analysis is based on comparing the evolution of applications between the prime d’activité on the one hand and the RSA (a benefit for which there was no reform or informational shock) on the other. Following the announcement of the reform, the prime d’activité—which remained a relatively recent and still little-known scheme (in 2018, one in four people did not know of its existence)—was the subject of very wide media coverage. It was massively relayed in the press, on radio and television, with explanations of eligibility criteria and target groups. Survey data confirm that a large number of people discovered the existence of the scheme at that time, which led to a sharp increase in applications at the beginning of 2019.

These results therefore suggest that a significant share of non-take-up before the reform was explained by a simple information problem, in particular a lack of knowledge of the very existence of the benefit. That said, it is also necessary to recall the limits of this “natural experiment” approach: while one can identify the effects of the reform on a particular population—those who reacted to this media shock—one cannot necessarily generalize the results to all non-take-up cases. It is likely that even today, many eligible people do not apply for the prime d’activité, but my study does not make it possible to determine which obstacles play a role for those people.

9. What lessons does your research bring for the design of public policies aimed at improving access to social benefits? In what way can your results contribute to a broader reflection on the effectiveness of social policies and on how to better reach the targeted populations?

The fact that the media coverage of the 2019 prime d’activité reform had such a marked impact on take-up makes it possible to consider avenues for reform in other contexts, whether for other social policies or in other countries.
Traditionally, economists and policymakers consider that, faced with information problems, one of the most effective responses is to implement targeted information policies (for example, personalized letters sent to individuals identified as potentially eligible and informing them of their rights in a personalized way). These targeted approaches, however, raise a number of difficulties: how to correctly identify people in situations of non-take-up? How to ensure that the message is received, understood, and perceived as legitimate? The results of studies on the subject show that this kind of intervention does not seem to work among populations that are not, or are barely, in contact with the administration.

By contrast, by extrapolation, my results suggest that a less targeted intervention—such as a national information campaign—can, paradoxically, prove more effective in combating non-take-up linked to a lack of information. We observe that the people who began to apply for the benefit following the reform were disproportionately individuals who had never been in contact with the Caisse des Allocations Familiales (CAF), the organization in charge of managing this benefit. This tends to show that such broad information campaigns can reach audiences that targeted approaches do not always manage to reach. Moreover, when a message is widely disseminated, it can also circulate through network effects, between eligible and non-eligible people, thereby strengthening its impact.

Of course, these campaigns have a cost, often higher than that of targeted mailings. It is therefore essential to balance these different elements—cost, reach, effectiveness—when thinking about the best tools to mobilize when we want to improve the take-up of social benefits.

Thomas Monnier – PhD student in economics at CREST, École polytechnique, IP Paris

In September 2025, he will join Hitotsubashi University (Institute for Advanced Study) as an assistant professor.

In his research, he applies structural approaches (mainly urban economics) to questions in development economics.

His thesis, The Informality Trade-Off: Wages and Rural-Urban Migration in South Africa, supervised by the economist Benoît Schmutz-Bloch, deals with the role of informal labor and housing in South Africa.

More broadly, he works on migration, urbanization, climate change, misallocation, structural transformation, and place-based policies.

As part of his doctorate, he also worked as a consultant for the World Bank.

1. Why did you choose South Africa as a field of study? What makes it both comparable and distinct from other countries?
South Africa is a middle-income country, richer and more developed than other economies in sub-Saharan Africa and comparable to those of the East Asia–Pacific region. These major geographical areas are those where the increase in the urban population—fueled by rural exodus—will be the strongest over the next 25 years. Given that informality can play a buffer role in the context of overcrowded cities, it seemed important to me to focus on one of the major economies of the area, rather than on Latin America for example—where many studies on informality have been published (notably due to access to high-quality data). The relatively recent legacy of apartheid also makes South Africa an interesting case study in itself, with migratory flows from rural to urban areas high for this level of urbanization (about 60% of the population) and specific dynamics of informality.

2. What are the dynamics of urban informality in South Africa?
Informality refers to all activities operating on the margins of existing laws and regulations, in the absence of a contract or enforcement mechanism (not to be confused with the criminal economy). If we consider cities, from an economic point of view, as local markets for employment and housing, speaking of urban informality essentially amounts to speaking of undeclared jobs and slums. These conditions are often far from ideal for workers and do not always benefit businesses, but they can constitute an interesting fallback solution, even a springboard towards better conditions (see next question). For example, an informal job can serve as an alternative to unemployment and informal housing as an affordable location close to employment centers. In the case of South African cities, the share of informal labor (about 30%) is typically lower than in comparable countries (about 60%), but the share of so-called “structural” unemployment is also higher (about 35% versus 5%). The share of informal housing is comparable to East Asia for example (about 25%), but nearly half of these homes consist of backyard installations attached to formal units (“backyarding”), an emerging phenomenon that is still little studied. These specific dynamics call for the modeling of corresponding congestion mechanisms that could in time apply to other countries in the area of interest, particularly in the case of poorly supported urban growth.

3. Is the analysis of informal urban labor markets enriched by taking into account spatial dynamics, notably rural–urban migration? What role does geographic mobility play in the resilience of workers to urban policies?
That is the whole point of the first chapter of my thesis (my “job market paper”). At the local level, even if informal employment is typically preferred to unemployment at a given moment, it is not always clear that it is a preferable option in the long term (if it is more difficult to find a formal job afterwards, for example). However, given the better professional opportunities that cities offer compared to rural areas, the choice of informal urban employment remains relevant for rural migrants seeking to integrate into the labor market (in the absence of a formal position and if one considers that a job is necessary to justify the mobility costs induced by migration). The formal–informal and rural–urban labor markets therefore do not operate in watertight compartments, and one must expect significant general equilibrium effects in the analysis of urban policies. Thus, I show that the negative impact of a formalization policy (reinforced controls on informal firms) on urban incomes leads to a decrease in geographic mobility from rural to urban areas: just as informal employment can serve as a fallback solution in the face of formal firms seeking to lower wages, rural employment can enable urban workers to adapt when their wage prospects actually decline. It should be noted that the direction of this effect is not obvious a priori, since wages can also be pushed upwards by increased competition between formal firms: the net effect between these two forces is an empirical question that justifies the use of a theoretical model bringing external validity to the mechanisms at work in the South African case.

4. In what way can informality be considered a rational response in contexts of high labor-market frictions? And what are the specific forms of frictions on the South African labor market that push individuals towards informality?
In a sense, frictions in the formal market are the raison d’être of the informal market: if supply and demand were perfectly aligned in one, workers and firms would have no interest in the other. In the case of the labor market, these frictions cover mechanisms of imperfect information, monetary or psychological costs for recruitment and job search, limited social networks, etc. For example, it may be simpler (and less costly) for a worker to signal their skills or motivation to a close contact, who would be all the more inclined to hire them without a contract if their activity is small and not very profitable. In doing so, each weighs the opportunity cost represented by the lack of access to social benefits or the resources of the formal market (access to capital, security, etc.). In the case of South Africa, these frictions seem to operate insofar as the rate of formal employment is comparable to that of other emerging countries. On the other hand, it would seem that the informal market is also characterized by strong frictions, which would explain the high unemployment rate: existing studies point to spatial frictions specific to informal activities (transport costs, zoning, permits) and to pressure from formal firms in certain key sectors (hospitality, commerce, agriculture). These factors contribute to the strong market power of firms over wages which underpins the results of the first chapter.

5. Is informal housing a fallback solution in the face of the scarcity of formal housing?
Absolutely, and that is the subject of the second chapter of my thesis, which focuses on the city of Cape Town. My co-authors and I show that the increase in the urban population induces upward pressure on formal housing prices in the absence of changes in planning constraints. Consequently, the share of informal housing increases, and this in different ways depending on the policies implemented to deal with urban growth: introducing a limit to urban sprawl further restricts the supply of formal housing and increases the share of slums, while the construction of social housing allows part of the slum population to move towards “backyarding,” typically less unsanitary. In my third chapter, we show that taking into account flood risks to which informal households are particularly subject does not fundamentally change this dynamic, even if it induces a reallocation in space. On the other hand, unlike the first chapter, I do not study the prospects of transition from informal to formal housing, nor do I look at the extent to which this issue of access to housing (and therefore to employment) affects return migration choices: these are avenues that I keep for future work.

6. In what way do the results of your research illustrate the ambiguous and ambivalent role of the informal sector?
As you suggested, my results depict informality as a constrained choice that is fully rational. However, it must be kept in mind that this choice is rarely optimal and constitutes at best a medium-term solution while awaiting a change in these constraints. Indeed, informal firms are partly protected from competition from the formal sector (by “parasitizing” workers’ job search), tend to remain less productive, and do not contribute to the tax effort. Informal housing limits households’ consumption capacity and steers them towards certain types and centers of employment. And that is without mentioning other aspects that I do not study directly: insurance logics, wealth accumulation, life cycles and intergenerational transmission, neighborhood effects and segregation, etc.

7. In your opinion, what urban public policies would make it possible to exit informality? And in the absence of adequate development policies, what economic functions can informality fulfill in emerging countries to allow an increase in workers’ well-being?
In my first chapter, I show that although coercive policies can allow a reduction in informality and an increase in production, they are not necessarily desirable insofar as they induce a decrease in workers’ well-being—and it is likely that I would obtain a similar result in the following chapters. The impact on the urbanization rate can also have an indirect negative effect on production in the presence of agglomeration economies and structural change mechanisms. It is therefore a matter of replacing or combining these policies with incentives for formality. For the latter to be effective, they must tackle the factors that justify the existence of the informal sector in the first place: low skill levels of the labor force, lack of resources and infrastructure, poor access to public services, administrative burdens, poorly defined legal framework, imperfect competition, strategic choices… The question of the relative effectiveness of these different options, their financing, and their political feasibility remains open.

8. Did the subject of your thesis or the orientations of your research evolve since the beginning of your PhD?
Yes, I had an initial interest in urban economics—both for its applications to public policy issues and for its overlaps with other social sciences—but I did not have a clear angle (in terms of object or method) for the whole thesis. My master’s dissertation was a historical economics project on public education, but I abandoned it, because another PhD student carried out similar work and because I realized that the associated data-processing work did not appeal to me. I then worked on another empirical project on fiscal zoning policies in France, which I also abandoned for lack of convincing results. A chance collaboration with the World Bank led me to the study of informal housing in the case of the city of Cape Town. The subject interested me, and I developed a taste for the underlying theoretical modeling and associated estimation techniques. This project gave me the opportunity to reflect more broadly on the role of the informal sector in the cities of developing countries—notably why it plays a predominant role there compared to richer countries—keeping South Africa as a case study. If I had to draw a lesson from this process, it would be to remain curious about new approaches, attentive to the evolution of one’s tastes, and open to opportunities that arise!

9. How do you envisage the continuation of your professional career?
For now, I am going to work as an assistant professor in a university in Japan. I will therefore pursue my research at the intersection of urban economics and development economics, broadening my interest in questions of informality to more general questions of structural transformation. Being in Japan will undoubtedly also be an opportunity to broaden my geographical scope. In parallel, I would like to continue interacting with policymakers, as I had begun to do with the World Bank. I find that this is an enriching experience in two respects: on the one hand, it allows impact-oriented research related to a given project to be brought to completion; on the other hand, it allows iteration on the project itself or the emergence of new ideas by better identifying certain concerns, field realities, etc. Belonging to the academic world makes it possible to maintain great freedom and a certain independence, even if the duration of projects can sometimes be frustrating. As long as I have the feeling that I continue to train myself and that I manage to deliver interesting results, I think I will remain there in the more distant future.