Since 2008, school choice in Paris has an income based affirmative action component granting low-income students preferential treatment in high school admissions. This policy is implemented as part of a centralized school choice procedure that assigns students to public schools based on a version of the Gale-Shapley deferred acceptance mechanism. Students’ priorities are determined using a point system that takes into account students’ academic performance and their district of residence. Low-income students, representing approximately 20 percent of high school entrants, are awarded a large bonus which gives them full priority at all public high schools within their district. Using comprehensive administrative data on the period 2004-2015, we exploit the introduction of this bonus in 2008 as a natural experiment to investigate the effects of income-based affirmative action on the high school outcomes and college access of both targeted and non-targeted students.

Using a general equilibrium heterogeneous agent model featuring health production, we quantify the contribution of health price in explaining cross-country differences in health expenditures and health status. Considering other country-specific explanatory factors, U.S. health prices are estimated to be 25% higher than those of European countries. This price differential explains more than 60% of the difference in health expenditures and more than half of the difference in health status between Europe and the U.S. Despite its large impact at the aggregate level, these price differences increase the lifetime cost of living of Americans by two percentage points.

We study strategic interaction among countries in pharmaceutical price regulation resulting from innovation-related spillovers. In our theoretical model, regulators’ pricing decisions affect consumer surplus both directly and indirectly, via firms’ R&D policies. We characterise two types of equilibrium, depending on whether countries price at, or above, the minimum level the industry is willing to accept to serve the market. Our results suggest that the combination of these two equilibria may imply a U-shape relationship between market share and pharmaceutical prices. We find empirical support for this hypothesis, using price data for 83 cancer drugs in 23 OECD countries. Our results contribute to the ongoing academic debate about the relationship between market size and prices, as well as the current policy debate about the merits of using supranational procurement policies to lower prices. In particular, we show that joint procurement can lead to lower or higher prices according to the sizes of the domestic markets which join together to create a single purchasing authority.

In 2020, New Zealand held referenda on legalization of recreational cannabis and assisted dying. Exploiting a panel survey, we estimate the effect of the signal provided by the referendum outcome on factual beliefs, and personal and social norms. Consistent with a model of norm-related beliefs, we find that the outcome of the cannabis referendum is associated with respondents updating their prior beliefs, not only about the social acceptability of recreational cannabis use, but also about the frequency of use and associated health risks, as well as changing personal moral views.

We study two-way-fixed-effects regressions with several treatment variables. Under a parallel trends assumption, we show that the coefficient on each treatment identifies the sum of two terms. The first term is a weighted sum of the effect of that treatment in each group and period, with weights that may be negative and sum to one. The second term is a sum of the effects of the other treatments, with weights summing to zero. Therefore, those coefficients are not robust to heterogeneous effects and may be contaminated by other treatments’ effects. We propose an alternative difference-in-differences estimator, robust to heterogeneous effects and immune to the contamination problem. To estimate, say, the first treatment’s effect, our estimator compares the outcome evolution of a group whose first treatment changes while its other treatments remain unchanged, to control groups whose treatments all remain unchanged, and with the same baseline treatments as the switching group.

In a broad class of discrete-time rational-expectations models, I consider generic simple instrument rules for stabilization policy, parameterized by a coefficient ϕ ∈ R and a time horizon h ∈ Z. Using two complex-analysis theorems, I establish some simple, necessary or sufficient conditions on ϕ and h for these rules to ensure local-equilibrium determinacy. The results have direct, general implications for the validity of the Taylor principle as a determinacy condition in monetary-policy models and other models, and for the ability of backward- and forward-looking stabilization policy to ensure determinacy. They also provide guidelines for finding rules with robust determinacy properties across alternative models.

We provide direct evidence of the asymmetry of information between local and foreign agents by looking at survey expectations. Using individual GDP growth and inflation forecasts by professional forecasters for a panel of emerging and advanced economies, we show that foreign forecasters make more mistakes than local forecasters. This finding is consistent across country and forecaster types and across sub-periods. This "home bias" in expectations is not due to a more irrational expectation formation by foreigners, but to local forecasters’ more precise information. Besides, consistent with a standard noisy information model where foreigners and locals observe the same public signals but locals have more precise private signals, foreign forecasts react more to public signals than the local ones. On the methodological side, we provide tests that identify differences in information frictions across groups.

Censorship makes new ideas less available to others, but also reduces the share of people choosing to develop non-compliant ideas. We propose a new method to measure the effect of censorship on knowledge growth, accounting for the endogenous selection of agents into compliant vs. non-compliant ideas. We apply our method to the Catholic Church’s censorship of books written by members of Italian universities and academies over the period 1400-1750. We highlight two new facts : once censorship was introduced, censored authors were of better quality than the non-censored authors, but this gap shrank over time, and the intensity of censorship decreased over time. These facts are used to identify the deep parameters of a novel endogenous growth model linking censorship to knowledge diffusion and occupational choice. We conclude that censorship reduced by 40% the average log publication per scholar in Italy, while adverse macroeoconomic processes are responsible for another 11% reduction. Interestingly, the induced reallocation of talents towards compliant activities explains half the effect of censorship.

Confronted with political opposition to the implementation of efficient direct carbon pricing, climate policy relies on alternative policy interventions, such as subsidies to renewables. This paper uses a dynamic macroeconomic model under a carbon budget to study climate policies constrained to keeping a constant level of the carbon tax.
We find that it is possible to implement the optimal trajectory by combining an increasing tax on electricity consumption with a feed-in-premium paid to electricity produced from renewable sources. Otherwise, when the climate policy relies on the second instrument only, the subsidy to renewables should be so large to foster rapid build up of specialized capital, that it would imply large investment costs and financial burden on the public budget, unless the carbon tax level could be initially set at a high level. Unfortunately, the two solutions with no or low welfare losses raise concerns on their political acceptability too.

We examine the extent to which programs offering financial rewards for losing weight should be adapted to participants’ characteristics. Specifically, we determine whether demanding and highly remunerative programs or – conversely – easily achievable but less remunerative programs should be proposed. We show that financial incentives are more likely to encourage efforts to lose weight if the level of difficulty of the objective and the reward in case of success increase with : 1) the degree of risk seeking ; 2) the degree of cross-prudence in health. This suggests that a personalized program based on preferences towards risk and correlation should be proposed to each participant.

We experimentally study an environment in which success requires a sufficient total contribution by members of a group. There are significant, and badly defined (ambiguous) uncertainties surrounding the chance of success and the total effort required to obtain it. We have a large representative sample of the Spanish population (1500 participants). A theoretical model with min-max preferences towards ambiguity and no strategic uncertainty predicts higher contributions in the ambiguous environments. However, we find that the ATE of ambiguity on contributions is zero (and the sample is large enough to have high power). We also find that both risk and ambiguity aversion significantly decrease contributions (and both dimensions of preferences are correlated, although not perfectly so). There are no significant interaction effects with the personal characteristics of the participant.

We propose a unifying representation of all normalized inequality indices, regardless of their properties. The key concept behind our results is that of negative extreme transfers (NETs), which are transfers from the poorest individuals to the richest individuals. This concept alone is rich enough to describe the entire space of income distributions. Indeed, our first result (Lemma 1) is that any income distribution can be obtained as an expansion from the equal distribution by applying a sequence of NETs. In other words, NETs constitute a mathematical basis of the space of income distributions. Our main representation theorem (Theorem 1) builds upon the NET decomposition of income distributions to describe any given inequality index based on the weight it attaches to all possible NETs. Accordingly, desirable properties (axioms) of inequality indices translate into properties of the weight function that are easy to check. We express well-known inequality indices according to this decomposition. We define the λ -NET Criterion, the requirement that the weighting function is greater than some parameter λ >0 . The stringency of the λ -NET Criterion filters out inequality indices as λ increases, so that one can therefore rank inequality measures on the basis of the λ -NET Criterion they satisfy. According to this ranking, we find that the Pietra index is the unique measure to satisfy the maximal λ.

The United States spends twice as much per person on pharmaceuticals as European countries, in large part because prices are much higher in the US. This fact has led policymakers to consider legislation for price controls. This paper assesses the effects of a US reference pricing policy that would cap prices in US markets by those offered in reference countries as proposed in the H.R.3 Lower Drug Costs Now Act of 2019. We estimate a structural model of demand and supply for pharmaceuticals in the US and reference countries like Canada where prices are set through a negotiation process between pharmaceutical companies and the government. We then simulate the counterfactual international reference pricing equilibrium, allowing firms to internalize the cross-country externalities introduced by this policy. We find that such a policy results in a slight decrease in US prices and a substantial increase in reference countries prices. The magnitude of these effects depends on the number, size and market structure of references countries. Overall, we find modest consumer welfare gains in the US but substantial losses in reference countries suggesting that this policy may not be the best way to introduce price controls in the US.

Models of statistical discrimination typically assume that employers make rational inference from (education) signals. However, there is a large amount of evidence showing that most people do not update their beliefs rationally. We use a model and two experiments to show that employers who are conservative, in the sense of signal neglect, discriminate more against disadvantaged groups than Bayesian employers. We find that such irrational statistical discrimination deters high-ability workers from disadvantaged groups from pursuing education, further exacerbating initial group inequalities. Excess discrimination caused by employer conservatism is especially important when signals are very informative. Out of the overall hiring gap in our data, around 40% can be attributed to rational statistical discrimination, a further 40% is due to irrational statistical discrimination, and the remaining 20% is unexplained or potentially taste-based.