Séminaires de recherche


A Simple Solution to Invalid Inference in the Random Coefficients Logit Model

Mardi | 2017-06-27
16h00-17h20 en salle des thèses

phillipp KETZ

We show that the standard approach to inference in the random coefficients logit model suffers from size distortions. The problem arises due to boundary issues and is aggravated when the model is parameterized with respect to standard deviations, which constitutes common practice. For example, in case of a single random coefficient, the asymptotic size of the nominal 95% confidence interval, which is obtained by inverting the two-sided t-test for the standard deviation, equals 83.65%. In seeming contradiction, we also find that standard errors can be unreasonably large. The proposed solution is to perform inference with respect to variances rather than standard deviations. This alleviates the problem of unreasonably large standard errors and, in combination with the estimator proposed in Ketz (2017a), allows the construction of confidence intervals that (uniformly) control asymptotic size.

The Persistent and the Transitory in Corporate Leverage Ratios

Mardi | 2017-06-13
16h00-17h20 en salle des thèses


We investigate the dynamics of observed and target leverage ratios and deviations from the targets. The cross-sectional persistence in observed leverage ratios is driven by highly persistent targets, whereas the time series variation is driven by transitory deviations from targets. Deviations are less persistent for firms that are over-levered and firms that are smaller, younger, focused, or have lower credit ratings. In recessions, excess leverage is less persistent for larger firms and is more persistent for smaller firms. Thus, consistent with dynamic trade-off theories, persistence is higher when the costs of deviating from targets are likely to be lower and when the costs of adjustment are likely to be higher.


Mardi | 2017-05-30
16h00-17h20 en salle des thèses

Olivier CARDI – Peter CLAEYS – Romain RESTOUT

This paper develops a two-sector open economy model with imperfect mobility of labor across sectors in order to account for time-series evidence on the aggregate and sectoral effects of a government spending shock. Using a panel of sixteen OECD countries over the period 1970-2007, our VAR evidence shows that a rise in government consumption i) increases hours worked and GDP and produces a simultaneous decline in investment and the current account, ii) increases non traded output relative to GDP and thus its output share (in real terms) and lowers the output share of tradables, and iii) causes both the relative price and the relative wage of non tradables to appreciate. While the second set of findings reveals that the government spending shock is biased toward non tradables and triggers a shift of resources for this sector, the third finding indicates the presence of labor mobility costs, thus preventing wage equalization across sectors. Turning to cross-country differences, empirically we detect a positive relationship between the magnitude of impact responses of sectoral output shares and the degree of labor mobility across sectors. Our quantitative analysis shows that our empirical findings for aggregate and sectoral variables can be rationalized as long as we allow for a difficulty in reallocating labor across sectors along with adjustment costs to capital accumulation. Finally, the model is able to generate a cross-country relationship between the degree of labor mobility and the responses of sectoral output shares which is similar to that in the data

Impulse Response Estimation By Smooth Local Projections

Mardi | 2017-05-23
16h00-17h20 salle des thèses


Vector Autoregressions (VAR) and Local Projections (LP) are well established methodologies for the estimation of Impulse Responses (IR). These techniques have complementary features: The VAR approach is more efficient when the model is correctly specified whereas the LP approach is less efficient but more robust to model misspecification. We propose a novel IR estimation methodology – Smooth Local Projections (SLP) – to strike a balance between these approaches. SLP consists in estimating LP under the assumption that the IR is a smooth function of the forecast horizon. Inference is carried out using semi-parametric techniques based on Penalized B-splines, which are straightforward to implement in practice. SLP preserves the flexibility of standard LP and at the same time can increase precision substantially. A simulation study shows the large gains in IR estimation accuracy of SLP over LP. We show how SLP may be used with common identification schemes such as timing restrictions and instrumental variables to directly recover structural IRs. We illustrate our technique by studying the effects of monetary shocks.

Fuzzy Differences-in-Differences

Mardi | 2017-05-16
16h00-17h20 salle des thèses


Difference-in-differences (DID) is a method to evaluate the effect of a treatment. In its basic version, a  » control group » is untreated at two dates, whereas a  » treatment group » becomes fully treated at the second date. However, in many applications of the DID method, the treatment rate only increases more in the treatment group. In such fuzzy designs, a popular estimator of treatment effects is the DID of the outcome divided by the DID of the treatment. We show that this ratio identifies a local average treatment effect only if two homogeneous treatment effect assumptions are satisfied. We then propose two alternative estimands that do not rely on any assumption on treatment effects, and that can be used when the treatment rate does not change over time in the control group. We prove that the corresponding estimators are asymptotically normal. Finally, we use our results to revisit Duflo (2001).


Mardi | 2017-05-09
16h00-17h20 salle des thèses


The integration of money in Ricardo’s theory of value and distribution is realised in Principles and the 1819-1823 papers by putting the standard of money centre stage. This raises two questions: Is money a commodity ? Is the standard of money (gold bullion) a commodity? The paper offers a negative answer to the first and a qualified positive one to the second. In Ricardo’s mature theory of money, in which the value of money depends on the value and on the price of the standard, money is neither a competitively-produced commodity nor a monopolised one. However, discarding a commodity-theory of money did not mean for Ricardo adopting a quantity theory: the determination of the value of money is sui generis in that the quantity of money is “regulated” by the standard. Although produced by the competition of capitals, this standard of money is a very special commodity: it should not be part of the system of production of commodities that determines the relative prices and the distribution of income. The fulfilment of this condition does not require the standard to be metallic: it could as well be a public bond purchased and sold for money by a central bank at a fixed price and traded on a secondary market – a situation not far from that of our modern economies.

Intermittent Discounting

Mardi | 2017-05-02
Salle des thèses de 16h00 à 17h20

A multi-period setting is studied in which a consumer incurs intermittent waiting costs until the consumption date. The way future utility is discounted satisfies (i) a preference for early consumption, but only during waiting periods when future consumption is reminded, and (ii) indifference between early and late consumption in non-reminding periods in which people are distracted away from future gratifications. With some additional hypotheses including exponential discounting of waiting costs, I show that the model replicates two important and robust features of intertemporal preferences: (i) decreasing impatience or hyperbolic discounting according to which the rate at which an outcome is discounted over time decreases as the time horizon gets longer and (ii) present bias i.e. the propensity to prefer immediate gratification to future ones. Intermittency of waiting states means that future utility is gradually but not regularly discounted with the passing of time. This slows down the discounting process and makes the consumer decreasingly impatient with the length of the delay.

Managing the Impact of Climate Change on Migration: Evidence from Mexico

Mardi | 2017-04-25
16h00-17h20 salle des thèses

Isabelle CHORT – Maëlys DE LA RUPELLE

This paper uses state-level migration flow data between Mexico and the U.S. from 1999 to 2011 to investigate the migration response to climate shocks and the mitigating impact of an agricultural cash-transfer program (PROCAMPO) and a disaster fund (Fonden). Our results suggest that droughts increase undocumented migration. Fonden amounts are found to mitigate the effect of climate shocks by lowering the undocumented migration response to precipitation anomalies. Similarly an increase in the share of PROCAMPO funds to the ejido sector decreases undocumented migration after a shock. By contrast, we find no robust evidence of a mitigating impact on documented migration.

Competition and credit procyclicality in European banking

Mardi | 2017-04-11
16h00-17h20 salle des thèses

Yannick LUCOTTE – Aurélien LEROY

This paper empirically assesses the effects of competition in the financial sector on credit procyclicality by estimating both an interacted panel VAR (IPVAR) model using macroeconomic data and a single-equation model with bank-level European banking data. The findings of these two empirical approaches highlight that an exogenous deviation of actual GDP from potential GDP leads to greater credit fluctuation in economies where (i) competition among banks and (ii) competition from non-bank financial institutions or direct finance (proxied by the financial structure) are weak. According to the financial accelerator theory, if lower competition strengthens the cyclical behavior of financial intermediaries, it follows that these  » endogenous developments in credit markets work to amplify and propagate shocks to the macroeconomy » (Bernanke et al., 1999). Furthermore, since credit booms are closely associated with future financial crises (Laeven and Valencia, 2012), our results can also be read as evidence that greater competition in the financial sphere reduces financial instability, which is in line with the competition-stability view denying the existence of a trade-off between competition and stability.

Cheap Talk ? Financial Sanctions and Non-Financial Activity

Mardi | 2017-04-04
Salle des thèses de 16h00 à 17h20


Sanctions restrict cross-border interactions and, therefore, not only put political and economic pressure on the target country, but they also adversely affect the sender country. This paper examines the effect of financial sanctions on the country imposing them. In particular, we analyze the business responses of German non-financial entities to the imposition of sanctions, using highly disaggregated, monthly data from the German balance of payments statistic for the period from 1999 through 2014 to identify firms affected by sanctions. Examining evidence from sanctions on 24 countries, we find, in a differences-in-differences setting, a measurable decline in German financial activities with the sanctioned country. For non-financial firms, however, the effect has been mainly on the extensive margin, while the volume of direct financial flows to and from these countries is much less affected. Moreover, firms doing business with sanctioned countries tend to be disproportionately large, making them largely immune to the reduction in business opportunities with selected partners. In fact, we find no effect of sanctions on firm-level variables such as employment or sales. Finally, sanctions imposed by the European Union alone, and therefore only enforced by their member countries instead of the United Nations, turn out to be evaded as flows with major trading partners of sanctioned countries increase. In sum, these results suggest that, despite the reduction in cross-border financial activities, the economic costs of financial sanctions to the sender country are limited.