In this talk, I consider the role of gender in constructing the knowledge produced in economics. I discuss the “gendering” of a discipline, in terms of how gender norms affect the accepted methods, research areas, data, and pedagogical approaches employed in the field. I apply this thinking to the discipline of economics and study how the gendered nature of the field makes the knowledge created in the discipline partial.

This paper investigates the role played by self-confidence in college applications. After documenting how much gender and social differences in self-confidence explain inequalities in access to college, we show that correcting underconfidence helps to close the gap in college applications. We collect unique data on student self-confidence using experiments on 3,000 French college applicants. We match this data with administrative data on real college applications. After measuring student under- (or over-) confidence in their academic abilities, we randomly corrected under- (or over-) confidence by providing some students information on their real rank in the grade distribution. Our results show that the best female students and the best students from low socio-economic status systematically underestimate their rank in the grade distribution. As a result, they apply to less selective programs. Correcting miscalibrated beliefs by providing information on the correct position in the grade distribution made the best students apply to more ambitious programs with stronger effects for female and low-SES students.

We conduct a large-scale online experiment, where we compare the short- and longer-term effects of fact checking to a brief media literacy intervention (ten tips to spot false news) as a means to debunk “fake news". We demonstrate that the effect of fact checking is limited to the specific fake news that are being targeted, whereas the media literacy intervention helps users to distinguish between false and correct information more generally, both immediately and two weeks after the intervention. Our results promote media literacy as an effective tool against “fake news", that is furthermore cheap, scalable, and easy-to-implement by social media platforms.

This paper investigates the contribution of managers to gender gaps and analyzes whether the over-representation of men in management positions puts women at a disadvantage. Relying on personnel data from one of the largest European manufacturing firms, we separate out the factors ex-plaining gender gaps. Adjusted pay gaps are positive, which means that men earn more than observationally equivalent women. A significant share of pay gaps can be explained by the sorting of men and women to different managers. More importantly, gender gaps in bonus payments causally depend on the manager’s gender. Accounting for worker and manager heterogeneity, bonus gaps are larger when the manager is male. This is driven by the fact that performance ratings are more favorable to men if handed out by a male manager. We present suggestive evidence that the relevance of manager gender for pay gaps is driven by discrimination rather than same-gender complementarities in productivity. However, independent of the root cause of these differences in evaluations by manager gender, the findings imply that a lower number of female managers increases gender gaps and thus constitutes a structural disadvantage for women.

We study the effect of exposure to the suffrage movement in England on women’s mobilisation into politics as electors and as politicians. We take advantage of the 1913 Women’s Suffrage Pilgrimage as a natural experiment to identify how exposure to suffrage affected women’s political mobilisation in England. We construct a novel micro-level dataset of geocoded political engagement from archival records. With these data, we compare parishes depending on their proximity to the Pilgrimage to test whether the women’s suffrage movement effectively mobilised newly enfranchised women to vote or to become politicians. Our approach challenges the conventional assumption that suffrage is a sufficient condition for the political inclusion of marginalised groups. We emphasise the importance of social movements to catalyse people’s actual practice of their political rights.

I provide new empirical evidence on the direct and indirect impact of services offshoring on local labour markets outcomes. Differently from the rest of the literature, I account for the effects that offshoring of services has on firms that are not directly involved in offshoring, but that are located in the same local labour market from where offshoring originates. I employ a unique detailed dataset on firms in the UK for the period 2000-2015 and exploits geographical differences in services trade flows across the country. Services offshoring is measured using precise information on firms’ location and trade in services flows. The analysis proceeds first estimating the causal average impact of services offshoring on local labour markets, showing positive elasticity of employment and wages. On a second step, it exploits firms’ heterogeneity and shows that part of the increase in average employment and wages is driven by the spillover effects of services offshoring on the firms not involved in services trade directly. Further, services offshoring widens the dispersion of firms, also when accounting for changes in the composition of firms through time and firms’ trade status. Finally, I look at the heterogeneous impact of services offshoring based on workers’ characteristics, showing higher elasticities of employment and wages to services offshoring of workers in managerial and professional occupation or with a higher level of education.

Since the economic and financial crisis of 2008/09, the world has seen a surge in the discomfort of electorates around the world, which materialized in changing political majorities and less predictable decisions flowing from direct democracy. Economists and political scientists concur that a change in the perceived income and wealth inequality in conjunction with an increased competition for resources may be part of the equation explaining this phenomenon. Beyond changes in consumer prices, are disposable, after-tax nominal incomes are of great importance to households' perception of inequality. While data on assets, incomes, and their distribution prior to taxation become more available, the profession has lacked a holistic, global view on how gross earnings are mapped into disposable income through labor income taxation and how this mapping has changed over past decades. The present paper aims to fill this gap by introducing and describing a data set of unprecedented size and detail covering personal labor income tax calculators and statutory tax codes for 251 countries and territories for the years 1980-2012. The tax calculators provide data on effective average and effective marginal tax rates for archetypal household varieties and allow for adjustment in the earnings status of two spouses, account for the deductions, credits and allowances that accrue to workers, spouses, and their children. In addition to presenting an overview of this data, the paper explores the impact of standard political factors on different measures of labor income tax measures, with particular attention on the built-in progressivity of labor income taxes.

This paper estimates the price and quantity effects of a new policy tool aimed at alleviating inflation in Argentina: a large, temporary VAT cut on food essentials coupled with a pre-existing price monitoring information system and price controls that dictate how VAT changes are shared between firms and consumers. Our findings show that governments can fine-tune VAT pass-through to prices with other complementary tools: VAT cuts combined with price controls/monitoring in grocery stores are substantially more effective at reducing prices than VAT cuts alone. This experience suggests that tax incidence is not an immutable parameter and it highlights the importance of enforcement when mandating incidence and stimulating demand.

Labor markets in low-income countries differ from those in high-income countries in ways related to labor market power. On the firms’ side, large entry barriers increase labor market concentration. On the workers’ side, workers find a readily accessible outside option in the self-employment sector. We study the scope, determinants, and implications of labor market power in low-income countries and its relation to self-employment opportunities. Using data on Peru, we document variation in employer concentration and high self-employment rates across local labor markets. Where employer concentration is high, self-employment rates are high, while earnings and average education are low for both wage and self-employed workers. To interpret these facts, we build a general equilibrium model where labor market power results from (i) barriers to firm entry and (ii) sorting of heterogeneous workers across wage work and self-employment. We structurally estimate the model to gauge the role of these fundamental forces for labor market power and study the effect of different policies promoting competition in the labor market. We find that the aggregate and distributional effects of these policies depend critically on whether they target the demand or supply side of the labor market.

I experimentally study how the transition to online education affects student performance, study efforts, social interactions and evaluations of teaching. Students are randomized students into live and online attendance of university lectures on a rotating attendance schedule.

In the United States, state and local governments compete to attract firms by offering discretionary subsidies. I use a private value English auction to model the subsidy bidding process and quantify the welfare effects of competition. The allocation of rents between states and firms depends on the heterogeneity in states' valuations for firms and the substitutability of locations. I find that competition increases welfare by about 10% over a subsidy ban, but states compete away the surplus, transferring the majority of rents to firms. Moreover, political concerns affect states’ valuations for firms. These findings dampen any interpretation of subsidy competition as an effective place-based policy.

Content selection algorithms of social media platforms have a reputation for creating filter bubbles, where users are primarily exposed to politically like-minded information. Taking a supply-side perspective, this study investigates the role of Facebook’s news feed algorithm for content created by media companies. Specifically, we evaluate the effects of two major changes in the algorithm on the political bias of stories posted by German news outlets. In December 2013 and August 2014, Facebook publicly announced measures designed to boost the dissemination of quality news stories. Using a differences-in-differences strategy, our results indicate greater levels of within-outlet ideological diversity of news posted on Facebook after the second algorithm update, relative to the print coverage. The larger part of the effect can be explained by a shift of the outlets towards more hard news and less trivia, which reduced deviations from the distribution of ideologically relevant expressions used in their print editions. The smaller part of the increase in outlets’ diversity can be attributed to changes in editorial policies that were likely motivated by efforts to cater to broader audiences.

Using administrative data for Upper Austrian workers from 2003-2013, we show that an extension in unemployment insurance (UI) duration increases unemployment length and impacts worker physical and mental health. These effects vary by gender. Specifically, we find that women eligible for an additional 9 weeks of UI benefits fill fewer opioid and antidepressant prescriptions and experience a lower likelihood of filing a disability claim, as compared to non-eligible unemployed women. Moreover, estimates indicate within-household spillovers for young children. For men, we find that extending UI benefit duration increases the likelihood of a cardiac event and eventual disability retirement filing.

This paper investigates the role of biased health perceptions as driving forces of risky health behavior. We define absolute and relative health perception biases, illustrate their measurement in surveys and provide evidence on their relevance. Next, we decompose the theoretical effect into its extensive and intensive margin: When the extensive margin dominates, people (wrongly) believe they are healthy enough to “afford” unhealthy behavior. Finally, using three population surveys, we provide robust empirical evidence that respondents who overestimate their health are less likely to exercise and sleep enough, but more likely to eat unhealthily and drink alcohol daily.

Modern public policy employs behavioral instruments and outcome evaluation to ensure the efficient use of public funds and increased welfare. Pfeifer et al. (2018) showed that banning smoking from school grounds effectively reduces smoking behavior. However, little is known about overall welfare effects of such interventions. In this paper, we conduct a more holistic evaluation and take a second look at heterogeneous effects of smoking bans on outcomes beyond smoking per se. We find that school smoking bans increase overall subjective well-being. Additionally, we propose an approach to identify smokers that are unlikely to reduce their smoking behavior due to the ban and show that those individuals suffer in terms of their well-being.

Case studies of cartels and recent theory suggest that repeated communication is key for stable cooperation under imperfect monitoring, where actions can only be observed with noise. We report the results of a laboratory experiment on cooperation under different monitoring and communication structures, which confirm this. Pre-play communication boosts cooperation at the beginning of an interaction under all monitoring structures. However, repeated communication is necessary to maintain cooperation under imperfect monitoring, where bad signals about others' actions can also occur after mutual cooperation. Subjects lack the foresight to discuss their responses to bad signals beforehand and thus need to renegotiate.

In this paper we construct a general estimable model of the technology underlying birth weight production that can be used to investigate the importance of different types of health inputs to the formation of birth weight. We draw on individual inpatient and outpatient records of the population in the Swedish province of Sk�ne providing us with detailed information on each health service encounter for a 10-year period. In addition, we merge the Swedish population censuses as well as tax and benefit records. As a result, we obtain the entire health and socioeconomic history of the mother prior to conception up to child birth. We then apply the methodology developed in Cunha et al. (2010) and Agostinelli and Wiswall (2018) to understand the productivity of prenatal care inputs at different points of pregnancy. The economic model considers a foetus in three trimesters of in utero development. A child's health at birth is produced by a (non)linear production technology, which has the interpretation of a skill multiplier and explains how medical inputs between the trimesters are linked. We refer to it as the health input multiplier in the production of birth weight. Such a production function not only allows the productivity of early investments in the first trimester to directly boost child health at birth, but via the health sector input multiplier it also raises the productivity of second trimester inputs. This channel has been implicitly ignored in the existing literature that examines birth weight production. Our preliminary results from estimating a (log) linear production technology suggests that child health is self-reinforcing, and trimester inputs exhibit a dynamic substitutability. Gaining an estimate of the elasticity of substitution parameter is a crucial piece of evidence for policymakers, allowing to identify the most critical period for subsequent development. With these structural estimates, we plan to conduct a series of policy simulations. These experiments will allow us to determine for different subgroups of mothers defined on the basis of their predetermined characteristics (e.g. age, income, education), which pathway of health sector investments would generate the healthiest in utero development.

How do development projects influence the geographic distribution of economic activity within low-income and middle-income countries? Existing research focuses on the effects of Western development projects on inter-personal inequality and inequality across different subnational regions. However, China has recently become a major financier of economic infrastructure in Africa, Asia, Latin America, the Middle East, and Central and Eastern Europe, and it is unclear if these investments diffuse or concentrate economic activity. We introduce an original dataset of geo-located Chinese Government-financed projects in 138 countries between 2000 and 2014, and analyze the effects of these projects on the spatial distribution of economic activity within host countries. We find that Chinese development projects in general, and Chinese transportation projects in particular, reduce economic inequality within and between subnational localities. Our results suggest that Chinese investments in "connective infrastructure" produce positive economic spillovers that lead to a more equal distribution of economic activity in the localities where they are implemented.

Recent studies have proposed causal machine learning (CML) methods to estimate conditional average treatment effects (CATEs). In this study, I investigate whether CML methods add value compared to conventional CATE estimators by re-evaluating Connecticut's Jobs First welfare experiment. This experiment entails a mix of positive and negative work incentives. Previous studies show that it is hard to tackle the effect heterogeneity of Jobs First by means of CATEs. I report evidence that CML methods can provide support for the theoretical labor supply predictions. Furthermore, I document reasons why some conventional CATE estimators fail and discuss the limitations of CML methods.

In standard tax compliance models, tax withholding at source is irrelevant. In these models, tax compliance is determined by a combination of enforcement (via audits and penalties), social motives, and third-party reporting, which deters evasion by enabling the tax authority to verify self-reported liability. The fact that third parties may also withhold taxes at source - and the impact of withholding on compliance - has largely been ignored. Yet tax withholding is common around the world: withholding of the personal income tax by employers is almost universal, and withholding is also applied to firms, especially in low-income countries. We provide a simple framework to rationalize the use of tax withholding as a compliance mechanism and test its predictions using administrative data from Costa Rica. We find that doubling the tax withholding rate applied by credit-card companies increases sales tax remittance among treated firms by 39% and aggregate sales tax revenue by 8%, even though the statutory tax rate and third-party reporting requirements remain unchanged. The mechanisms are a default payment effect and a change in enforcement perceptions. Our findings contribute to a broader debate about the determinants of fiscal capacity among countries at different levels of economic development.

Long-term unemployment imposes large costs on both individuals and society. Reducing long-term unemployment therefore is of first-order importance for public policy. We analyze to what extent work meaning - the significance of a job for others or for society - increases the willingness to accept a job and job performance of employed and unemployed individuals. To this end, we conduct a large-scale online experiment with a representative sample of the population. All of our subjects participate in the "Panel Study of Labour Market and Social Security'' (PASS) of the German Labour Agency. In the experiment, we offer a job that can be completed from home. We elicit subjects' reservation wage for this job through the Becker-DeGroot-Marschak mechanism. The treatment variation is the description of the job as having either "high" or "low'' meaning. To study heterogeneity in subjects' reservation wages, we link their behaviour in the experiment to rich survey data provided by PASS.

Superstition is a widespread phenomenon. We show that believers of the Taiwanese Ghost Month adjust (due to false beliefs) their behavior in significant ways with consequences for their health. We document adaptions in risky behavior and in the demand for health-care utilization. The first effect leads to a reduction in mortality, mainly driven by a drop in accidental deaths. The second effect changes mainly, but not exclusively the timing of health-care utilization. We observe the avoidance of surgeries and births during the Ghost Month. Estimated effects are identified by idiosyncratic variation of the Ghost Month across Gregorian calendar years, and based on high-quality administrative data covering the whole population.

We present the first evidence on the role of occupational choices and acquired skills for migrant selection. Combining novel data from a representative Mexican task survey with rich individual-level worker data, we find that Mexican migrants to the United States have higher manual skills and lower cognitive skills than non-migrants. Results hold within narrowly defined region-industry-occupation cells and for all education levels. Consistent with a Roy/Borjas-type selection model, differential returns to occupational skills between the United States and Mexico explain the selection pattern. Occupational skills are more important to capture the economic motives for migration than previously used worker characteristics.

This paper explores the role of cheap excuses in product choice. If a product improves upon one ethically relevant dimension, agents may care less about other independent ethical facets of the product. Opting for a product that fulfills one ethical aspect may thus suffice for keeping a high moral self-image in agents, and render it easier to ignore other ethically relevant aspects they would otherwise care about. The use of such cheap excuses could thus lead to a 'static moral self-licensing' effect. This would extend the logic of the well-known moral self-licensing over time. Our experimental study provides empirical evidence that the static counterpart of moral self-licensing exists. Furthermore, effects spill over to unrelated, ethically relevant contexts later in time. Thus, static moral self-licensing and moral self-licensing over time can amplify each other. Outsiders, though monetarily incentivized for correct estimates, are completely oblivious to the effects of moral self- licensing, both, static and over time.

In this paper, we reconcile two seminal strands in the literature on export dynamics and economic development. A traditional argument has emphasized that the type of goods that countries export matters (Singer, 1949; Prebisch, 1950; Hausmann et al., 2007). A more recent literature has conversely highlighted the potential of vertical export upgrading by demonstrating that international division of labor takes place by high-income countries vertically specializing into higher-quality product varieties across sectors (Schott, 2004). We use export and import data of 4,800 products for 154 countries between the mid-1990s and mid-2000s to highlight a series of stylized facts. First, we complement previous findings of convergence in export unit values and export quality by demonstrating amazing uniformity of convergence speed across products and accelerated convergence for exporters at lower income levels. Second, we analyze which products have most potential for quality upgrading by analyzing the length of their quality ladders and the speed of unit value increases. Particularly, we analyze how those aspects relate to complexity across products. Finally, we switch to a country perspective and ask the question how their income level is associated with their export portfolio in terms of the analyzed product characteristics and discuss the tradeoffs that emerge between sticking to current comparative advantage and switching to high-potential export goods.

The representation of women in politics has increased across the globe over the last few decades. Accordingly, interest in whether this rise in female representation is merely symbolic or whether it has substantive implications for policy has grown as well. We investigate this question for the legislative branch of government using unique hand-collected data on some 300,000 candidates for open-list local councils elections in Germany. Besides the name and gender of the candidates, our data covers initial and final list ranks, number of votes, party affiliation, current employment as well as a number of further characteristics. To identify a causal effect, we use the fact that in our open-list setting, the gender of the candidate who wins a mixed-gender election for the last council seat that accrues to a party is quasi-random. Our results indicate that councils with more women invest more in "pro-female" policies, notably publicly provided child care facilities, confirming that female representation has substantive implications for policy choices. We also find that these policy choices have broader economic effects. In particular, municipalities with a higher share of women in the council witness stronger female employment growth (male employment also increase, but to a smaller extent). Regarding mechanisms, our results indicate that a higher representation in the council ostensibly makes it easier for women to translate their concerns into policy. In line with this interpretation, we do not observe that female legislators raise aggregate employment by expanding public employment. There is no evidence that the effect is driven by partisan considerations, such as left-wing legislators building strategic coalitions with (conservative) female legislators. We also do not find that female legislators are less (or more) effective when the mayor is male.

We study the persistent effect of initial labor market conditions for labor marketentrants in the United States from 1976 to today on earnings, receipt of governmentsupport, and mortality by education, gender, and race groups. We find that adverse ef-fects are larger for workers without a college degree and nonwhites. While these effectsare partly offset by increases in the receipt of food stamps for the least advantaged, wefind persistent increases in poverty. We also find moderate increases in mortality later inlife, suggesting an unlucky start still has adverse effects once earnings and wage losseshave faded.

The combination of occupational regulation at the local market level and labor mobility across local markets is common. We empirically study a labor market in which a district-specific entry examination is coupled with labor mobility across districts. The analysis exploits a change in the rules for grading entry examinations, which were first graded locally and then randomly assigned to a different district. We document that regulation leads to 1) extreme heterogeneity across markets in admission outcomes, 2) unfair (discriminatory) admission procedures, and 3) inefficient mobility of workers. The impact of the reform on exam outcomes and grading standards is consistent with strategic interaction between local licensing boards.

We study the dynamic causal response of local budgets to a shift in revenues. We first isolate quasi-experimental variation in fiscal transfers within Germany's municipal fiscal equalization scheme. Because transfers are unconditional and depend on population counts, the 1987 Census led to sudden and permanent shifts in revenues across municipalities. Using an event-study design, we then track the causal response of municipal budgets to revenue gains or losses. In contrast to prior work, we focus on the dynamics of this response over time. Local budgets do not adjust instantly. The total level of government spending adapts over three to four years, predominantly driven by capital and infrastructure investments. Local tax rates adjust even more slowly. Business and property tax rates decrease in response to revenue gains, but the decline stretches over at least one decade. We contrast these findings with prior evidence on the "flypaper effect" and other empirical regularities, as well as theoretical work on local public finance.

Innovation policy encourages firms' participation in co-funding schemes for R&D projects. Yet, the distribution of awarded grants tends to be highly skewed with few firms receiving the lion's share of the public money. From a welfare perspective, supporting a small number of selected firms can be an efficient policy provided that the subsidies trigger knowledge spillover generating R&D in the subsidized firms. This study investigates whether the effect of subsidies go beyond the subsidized firms by affecting productivity of non-subsidized ones. We find that research subsidies affect productivity of non-subsidized positively, while development subsidies do not. Dynamic model specifications further show that in the longer run, development subsidies lead to business stealing by promoting product and process development in the subsidized firms with little positive spillovers to the non-recipients. When distinguishing R&D-active and non-R&D-active firms among the non-subsidized firms, we see that adverse competitive effects are stronger for the latter. Conducting own R&D may therefore increase firms' ability to realize effective knowledge spillovers and to avoid being harmed by competitive pressure from other firms product or process innovations.

In this paper we investigate to what extent the childhood healthcare environment influences later life health outcomes. We examine a fundamental re-organisation of the healthcare environment in the U.K., which occurred through the introduction of the National Health Service (NHS) in July 1948. Immediate large decreases in infant mortality of 17% ensued, which were focused on the neo-natal period and larger for individuals who prior to the NHS had a lower access to medical services. We combine historic county-level data with the Office of National Statistics Longitudinal Study of linked census records combined with administrative mortality data, and a large new dataset recording health measurements linked to administrative hospital records to assess the long run impact of birth exposure to the NHS on health and mortality 50 to 60 years after its introduction. Our findings indicate that survival rates are systematically higher among lower class individuals whose post-natal care expanded through the NHS, with the magnitude of the effect increasing monotonically with age. We supplement these findings with analysis of hospital records, which reveal a similar decrease in hospitalisations for cardiovascular disease, one of the major causes of death, for lower class individuals.

We construct a novel consumption-saving model where households must infer their permanent income from income realizations together with an additional noisy private signal. We then study how consumption responds to income shocks in our framework where consumers cannot perfectly distinguish permanent from transitory income shocks. The degree of imperfect information has important consequences for the interpretation of transmission parameters to transitory and permanent income shocks. Our theoretical results motivate the identification of the degree of knowledge from panel data on income and consumption. Finally, we estimate a high degree of knowledge in the Panel Study of Income Dynamics.