Publications

Résumé In this paper, we investigate on 39 Variable Selection procedures to give an overview of the existing 1 literature for practitioners. "Let the data speak for themselves" has become the motto of many applied researchers 2 since the amount of data has significantly grew. Automatic model selection have been raised by the search 3 for data-driven theories for quite a long time now. However while great extensions have been made on the 4 theoretical side still basic procedures are used in most empirical work, eg. Stepwise Regression. Some reviews 5 are already available in the literature for variable selection, but always focus on a specific topic like linear 6 regression, groups of variables or smoothly varying coefficients. Here we provide a review of main methods and 7 state-of-the art extensions as well as a topology of them over a wide range of model structures (linear, grouped, 8 additive, partially linear and non-parametric). We provide explanations for which methods to use for different 9 model purposes and what are key differences among them. We also review two methods for improving variable 10 selection in the general sense. 11
Mots clés Sparse models, Variable selection, Automatic modelling
Publication HAL Google Scholar
Résumé This paper is intended to bridge the theoretical literature describing efficient intra-household behaviour and the development literature that collects empirical regularities pointing toward the existence of strategic decision-making among spouses. It examines the key elements of the collective model and discusses its relevance to analysing intra-household behaviour in poor countries. It explores the role that risk and uncertainty, information asymmetries, power imbalances, arranged marriages, strategic investment, gender norms, and extended households play in the attainment of efficiency.
Mots clés Household, Efficiency, Development, Strategic behaviour
HAL Google Scholar
Résumé The aim of the paper is to shed light on the question of why a country decides to set up a Sovereign Wealth Fund (SWF). Despite the recent financial crisis, 43 SWFs have been created between 2005 and 2014. In particular, we test if the emergence of these new recent funds can be explained by the following economic, political and institutional factors : i) the excess foreign exchange reserves due to natural resources rents or persistent current account surpluses ; ii) the volatility of commodity prices ; iii) a way to mitigate the "Dutch Disease" effect and iv) the governance of the country. We test these hypotheses on a sample of 37 countries that created a SWF over the period 2000-2014 and compare them to a large panel of countries that did not set up a SWF. In order to allow the temporal dimension as well as the unobserved heterogeneity between SWFs, a Logit panel model with random effects is estimated. The results show that countries for which the creation of a SWF is more appropriate are those with foreign exchange excess reserves, which are dependent on a commodity and on its volatility and which suffer from an appreciation of the real exchange rate. We also find that non-democratic countries with a high level of corruption are more likely to create a SWF. Our results may be of interest for policymakers debating whether or not it can be optimal for the country to establish a SWF. "Modern Sovereign Wealth Funds are not new. The first, the Kuwait Investment Office, was set up in 1953 just as Edmund Hillary and Tenzing Norgay were setting out to climb Mount Everest. The number of funds has been increasing since then like the traffic on the slopes of Everest" (John Gieve, former deputy Governor of Bank of England in a speech in London, 2008).
Mots clés G23, Country Factors, Natural Resources Rents, Sovereign Wealth Funds, F39, H59 Corresponding author, Logit Panel Model JEL classification E21, E61
Publication HAL Google Scholar
Résumé Le 3 e Colloque international Philosophie économique a été organisé les 15 et 16 juin 2016 à Aix-en-Provence par le Greqam/Amse. Il a réuni des conférenciers venus de vingt-quatre pays, avec pour orateurs pléniers Cristina Bicchieri, John Davis et Daniel Hausman autour du thème de « l’agent économique et ses représentations ». Nous avons rassemblé dans le présent numéro spécial une sélection de textes présentés et discutés à cette occasion. La manière dont on peut représenter « l’agent économique » importe à la fois du point de vue de la philosophie économique qui l’interroge, de la théorie économique qui l’utilise, et des travaux empiriques qui s’appuient sur les modèles qu’elle propose et, parfois, impose. Elle importe encore afin de comprendre la façon dont les institutions émergent, dont les sociétés s’organisent et dont divers mécanismes contribuent au « bien-être général » – que ceux-ci apparaissent spontanément ou qu’ils soient le résultat d’une création volontaire et pragmatique. Elle importe enfin également pour envisager (et possiblement) corriger des situations dans lesquelles les marchés sont incomplets ou n’existent tout simplement pas. La question de l’agence (ou de l’« agentivité ») se pose en économie d’une manière dont la spécificité mérite discussion – prolongeant ainsi les travaux sur cette question. Malgré la diversité des termes rencontrés, en anglais comme en français (le colloque s’étant tenu dans les deux langues), pour désigner cet élément-clé des analyses proposées par les diverses sciences sociales, il semble que les économistes lui prêtent un intérêt qui ne se retrouve pas autant dans les autres disciplines…
HAL Google Scholar
Résumé Since the early 2000s, the question of access to medicines at affordable prices for Southern populations has appeared as one of the major challenges for the international governance of health. But what is at stake is the creation of market for medicines in the global South, particularly countries in the Sub-Saharan Africa. These markets are new in nature in the sense that they are driven by international organisations where Southern firms, especially from India, occupy an increasingly important position. However, the specificity of these markets and the way they are constituted have been little analysed. In this article, we suggest focusing the attention on the constitution of the market of antimalarial drugs and highlighting the role played by Southern firms in this market. Our study focuses on the public sector market of antimalarial drugs. We provide an institutional and quantitative analysis of the creation of this public market. We then discuss the growing importance of the Southern firms, mainly Indian and Chinese, in this market.
Publication HAL Google Scholar
Résumé Background: Measuring homophobia at country level is important to guide public health policy as reductions in stigma are associated with improved health outcomes among gay men and other men who have sex with men. Methods: We developed a Homophobic Climate Index incorporating institutional and social components of homophobia. Institutional homophobia was based on the level of enforcement of laws that criminalise, protect or recognise same-sex relations. Social homophobia was based on the level of acceptance and justifiability of homosexuality. We estimated the Index for 158 countries and assessed its robustness and validity. Results: Western Europe is the most inclusive region, followed by Latin America. Africa and the Middle East are home to the most homophobic countries with two exceptions: South Africa and Cabo Verde. We found that a 1% decrease in the level of homophobia is associated with a 10% increase in the gross domestic product per capita. Countries whose citizens face gender inequality, human rights abuses, low health expenditures and low life satisfaction are the ones with a higher homophobic climate. Moreover, a 10% increase in the level of homophobia at country level is associated with a 1.7-year loss in life expectancy for males. A higher level of homophobia is associated with increased AIDS-related death among HIV-positive men. Conclusion: The socioecological approach of this index demonstrates the negative social, economic and health consequences of homophobia in low- and middle-income countries. It provides sound evidence for public health policy in favour of the inclusion of sexual minorities.
Publication HAL Google Scholar
Résumé We provide an axiomatic characterization of a family of criteria for ranking completely uncertain and/or ambiguous decisions. A completely uncertain decision is described by the set of all its consequences (assumed to be finite). An ambiguous decision is described as a set of possible probability distributions over a set of prizes. Every criterion in the family compares sets on the basis of their conditional expected utility , for some “likelihood” function taking strictly positive values and some utility function both having the universe of alternatives as their domain.
Mots clés D81, Axioms JEL classification numbers D80, Ranking Sets, Expected Utility, Conditional Probabilities, Ambiguity, Ignorance
Publication HAL Google Scholar
Résumé Generic Bordeaux red wine (basic claret) can be regarded as being similar to an agricultural commodity. Production volumes are substantial, they are traded at high frequency and the quality of the product is relatively homogeneous. Unlike other commodities and the top-end wines (which represent only 3% of the traded volume), there is no futures market for generic Bordeaux wine. Reliable forecasts of prices can to large extent replace this information deficiency and improve the functioning of the market. We use state-space methods with monthly data to obtain a univariate forecasting model for the average price. The estimates highlight the stochastic trend and the seasonality present in the evolution of the price over the period 1999 to 2016. The model predicts the path of wine prices out of sample reasonably well, suggesting that this approach is useful for making reasonably accurate forecasts of future price movements.
Mots clés Q11, L66, Forecasting JEL CLASSIFICATION C53, State-space methods, Wine prices, Forecasting
Publication HAL Google Scholar
Résumé Heterogeneity in how some independent variables affect a dependent variable is pervasive in many phenomena. In this respect, this paper addresses the question of constant versus nonconstant effect through quantile regression modelling. For linear quantile regression under endogeneity, it is often believed that the fitted-value setting (i.e., replacing endogenous regressors with their exogenous fitted-values) implies constant effect (that is: the coefficients of the covariates do not depend on the considered quantile, except for the intercept). Here, it is shown that, under a weakened instrumental variable restriction, the fitted-value setting can allow for nonconstant effect, even though only the constant-effect coefficients of the model can be identified. An application to food demand estimation in 2012 Egypt shows the practical potential of this approach.
Mots clés Two-stage estimation Quantile regression Fitted-value setting Nonconstant effect Partial identification
Publication HAL Google Scholar
Résumé This article presents an assessment of individual uncertainty about longevity. A survey performed on 3,331 French people enables us to record several survival probabilities per individual. On this basis, we compute subjective life expectancies (SLE) and subjective uncertainty regarding longevity (SUL), the standard deviation of each individual’s subjective distribution of her or his own longevity. It is large and equal to more than 10 years for men and women. Its magnitude is comparable to the variability of longevity observed in life tables for individuals under 60, but it is smaller for those older than 60, which suggests use of private information by older respondents. Our econometric analysis confirms that individuals use private information—mainly their parents’ survival and longevity—to adjust their level of uncertainty. Finally, we find that SUL has a sizable impact, in addition to SLE, on risky behaviors: more uncertainty on longevity significantly decreases the probability of unhealthy lifestyles. Given that individual uncertainty about longevity affects prevention behavior, retirement decisions, and demand for long-term care insurance, these results have important implications for public policy concerning health care and retirement.
Publication HAL Google Scholar