Institut National de la Statistique et des Etudes Economiques

This article investigates estimation of finite population totals in the presence of univariate or multivariate auxiliary information. Estimation is equivalent to attaching weights to the survey data. We focus attention on the several weighting systems that can be associated with a given amount of auxiliary information and derive a weighting system with the aid of a distance measure and a set of calibration equations. We briefly mention an application to the case in which the information consists of known marginal counts in a two- or multi-way table, known as generalized raking. The general regression estimator (GREG) was conceived with multivariate auxiliary information in mind. Ordinarily, this estimator is justified by a regression relationship between the study variable y and the auxiliary vector x. But we note that the GREG can be derived by a different route by focusing instead on the weights. The ordinary sampling weights of the kth observation is 1/πk , where πk is the inclusion probability of k. We show that the weights implied by the GREG are as close as possible, according to a given distance measure, to the 1/πk while respecting side conditions called calibration equations. These state that the sample sum of the weighted auxiliary variable values must equal the known population total for that auxiliary variable. That is, the calibrated weights must give perfect estimates when applied to each auxiliary variable. That is a consistency check that appeals to many practitioners, because a strong correlation between the auxiliary variables and the study variable means that the weights that perform well for the auxiliary variable also should perform well for the study variable. The GREG uses the auxiliary information efficiently, so the estimates are precise; however, the individual weights are not always without reproach. For example, negative weights can occur, and in some applications this does not make sense. It is natural to seek the root of the dissatisfaction in the underlying distance measure. Consequently, we allow alternative distance measures that satisfy only a set of minimal requirements. Each distance measure leads, via the calibration equations, to a specific weighting system and thereby to a new estimator. These estimators form a family of calibration estimators. We show that the GREG is a first approximation to all other members of the family; all are asymptotically equivalent to the GREG, and the variance estimator already known for the GREG is recommended for use in any other member of the family. Numerical features of the weights and ease of computation become more than anything else the bases for choosing between the estimators. The reasoning is applied to calibration on known marginals of a two-way frequency table. Our family of distance measures leads in this case to a family of generalized raking procedures, of which classical raking ratio is one.

This paper studies the links between productivity, innovation and research at the firm level. We introduce three new features: (i) A structural model that explains productivity by innovation output, and innovation output by research investment: (ii) New data on French manufacturing firms, including the number of European patents and the percentage share of innovative sales, as well as firm-level demand pull and technology push indicators; (iii) Econometric methods which correct for selectivity and simultaneity biases and take into account the statistical features of the available data: only a small proportion of firms engage in research activities and/or apply for patents; productivity, innovation and research are endogenously determined; research investment and capital are truncated variables, patents are count data and innovative sales are interval data. We find that using the more widespread methods, and the more usual data and model specification, may lead to sensibly different estimates. We find in particular that simultaneity tends to interact with selectivity, and that both sources of biases must be taken into account together. However our main results are consistent with many of the stylized facts of the empirical literature. The probability of engaging in research (R&D) for a firm increases with its size (number of employees), its market share and diversification, and with the demand pull and technology push indicators. The research effort (R&D capital intensity) of a firm engaged in research increases with the same variables, except for size (its research capital being strictly proportional to size). The firm innovation output, as measured by patent numbers or innovative sales, rises with its research effort and with the demand pull and technology indicators, either directly or indirectly through their effects on research. Finally, firm productivity correlates positively with a higher innovation output, even when controlling for the skill composition of labor as well as for physical capital intensity.

This paper applies the standard rational expectations competitive storage model to the study of thirteen commodities. It explains the skewness, and the existence of rare but violent explosions in prices, coupled with a high degree of price autocorrelation in more normal times. A central feature of the model is the explicit recognition of the fact that it is impossible for the market as a whole to carry negative inventories, and this introduces an essential non-linearity which carries through into non-linearity of the predicted commodity price series. For most of the thirteen commodity prices, the behaviour of prices from one year to the next conforms to the predictions of the theory about conditional expectations and conditional variances. However, given the non-linearity both of the model and of the actual prices, such conformity is not enough to ensure that the theory yields a complete account of the data. In particular, the analysis does not yield a fully satisfactory explanation for the high autocorrelation observed in the data.

In this paper we present inference methods which are based on an ‘incorrect’ criterion, in the sense that the optimization of this criterion does not directly provide a consistent estimator of the parameter of interest. Moreover, the argument of the criterion, called the auxiliary parameter, may have a larger dimension than that of the parameter of interest. A second step, based on simulations, provides a consistent and asymptotically normal estimator of the parameter of interest. Various testing procedures are also proposed. The methods described in this paper only require that the model can be simulated, therefore they should be useful for models whose complexity rules out a direct approach. Various fields of applications are suggested (microeconometrics, finance, macroeconometrics).

The first goal of this paper is to provide a simple and general test of the presence of asymmetric information in contractual relationships within a competitive context. We also argue that insurance data are particularly well suited to such empirical investigations. To illustrate this claim, we use data on contracts and accidents to investigate the extent of asymmetric information in the French market for automobile insurance. Using various parametric and nonparametric methods, we find no evidence for the presence of asymmetric information in this market.

International audience

With a large international sample (n = 8317), the present study examined which beliefs and attitudes about COVID-19 predict 1) following government recommendations, 2) taking health precautions (including mask wearing, social distancing, handwashing, and staying at home), and 3) encouraging others to take health precautions. The results demonstrate the importance of believing that taking health precautions will be effective for avoiding COVID-19 and generally prioritizing one's health. These beliefs continued to be important predictors of health behaviors after controlling for demographic and personality variables. In contrast, we found that perceiving oneself as vulnerable to COVID-19, the perceived severity of catching COVID-19, and trust in government were of relatively little importance. We also found that women were somewhat more likely to engage in these health behaviors than men, but that age was generally unrelated to voluntary compliance behaviors. These findings may suggest avenues and dead ends for behavioral interventions during COVID-19 and beyond.

Access to private information is shown to generate both the incentives and the ability to manipulate asset markets through strategically distorted announcements. The fact that privileged information is noisy interferes with the public's attempts to learn whether such announcements are honest; it allows opportunistic individuals to manipulate prices repeatedly, without ever being fully found out. This leads us to extend Sobel's [1985] model of strategic communication to the case of noisy private signals. Our results show that when truthfulness is not easily verifiable, restrictions on trading by insiders may be needed to preserve the integrity of information embodied in prices.

This paper analyzes the relationship between the diffusion of new technologies and the decentralization of firms. Centralized control relies on the information of the principal, which we equate with publicly available information. Decentralized control, on the other hand, delegates authority to a manager with superior information. However, the manager can use his informational advantage to make choices that are not in the best interest of the principal. As the available public information about the specific technology increases, the tradeoff shifts in favor of centralization. We show that firms closer to the technological frontier, firms in more heterogeneous environments, and younger firms are more likely to choose decentralization. Using three data sets on French and British firms in the 1990s, we report robust correlations consistent with these predictions.

By buying cheap and selling dear, risk-neutral commodity speculators can smooth commodity prices and induce serial dependence in price even when none would exist under a simple process of supply and demand. Commodity prices are variable and strongly positively correlated from one year to the next. The variability is often explained by supply factors, and the autocorrelation by the activities of speculators. We show that this explanation is not consistent with the evidence. Speculation can substantially increase autocorrelation for prices that are weakly autocorrelated in its absence, but not to the high levels that are observed in the data.

We analyze a model of optimal consumption and portfolio selection in which consumption services are generated by holding a durable good. The durable good is illiquid in that a transaction cost must be paid when the good is sold. It is shown that optimal consumption is not a smooth function of wealth; it is optimal for the consumer to wait until a large change in wealth occurs before adjusting his consumption. As a consequence, the consumption based capital asset pricing model fails to hold. Nevertheless, it is shown that the standard, one factor, market portfolio based capital asset pricing model does hold in this environment. It is shown that the optimal durable level is characterized by three numbers (not random variables), say x, y, and z (where x

Egalement paru dans : Document de travail du LEA ; 00-02

We propose the name generalized raking for the class of procedures developed in this article, because the classical raking ratio of W. E. Deming is a special case. Generalized raking can be used for estimation in surveys with auxiliary information in the form of known marginal counts in a frequency table in two or more dimensions. An important property of the generalized raking weights is that they reproduce the known marginal counts when applied to the categorical variables that define the frequency table. Our starting point is a class of distance measures and a set of original weights in the form of the standard sampling weights 1/π k , where π k is the inclusion probability of element k. New weights are derived by minimizing the total distance between original weights and new weights. The article makes contributions in three areas: (1) statistical inference conditionally on estimated cell counts, (2) simple calculation of variance estimates for the generalized raking estimators, and (3) presentation of the new computer software CALMAR. Our conditional approach highlights the role played by interaction between the factors that define the frequency table. Absence of interaction implies that generalized raking is as efficient as complete post-stratification. The variance estimates we propose are calculated with the aid of the residuals from the fit of an additive analysis of variance (ANOVA) model. The CALMAR software, recently developed at I.N.S.E.E., is now used in various national surveys for calculating generalized raking weights. We illustrate its use with the aid of data from the 1990 survey of living conditions in France. In this application a table in seven dimensions with known marginal counts is used for generalized raking.

We review the econometric literature on measuring the returns to R&D.The theoretical frameworks that have been used are outlined, followed by an extensive discussion of measurement and econometric issues that arise when estimating the models.We then provide a series of tables summarizing the major results that have been obtained and conclude with a presentation of R&D spillover returns measurement.In general, the private returns to R&D are strongly positive and somewhat higher than those for ordinary capital, while the social returns are even higher, although variable and imprecisely measured in many cases.

The classical theory of commodity price determination integrates myopic supply and demand on the one hand with competitive storage (speculation) under rational expectations on the other. Taking into account the fact that inventories mist; be non-negative, this paper derives from the theory testable implications on the behavior of prices, and makes a first attempt to confront these implications with the empirical evidence. The nonlinearities turn out to be a crucial ingredient in matching the stylized facts, particularity the asymmetries and the sharp upward flares that characterize many commodity prices. The model, simple as it is, goes a long way in reproducing the main features of the data for a range of commodities.

The problem of underdetermination is thought to hold important lessons for philosophy of science. Yet, as Kyle Stanford has recently argued, typical treatments of the offer only restatements of familiar philosophical problems. Following suggestions in Duhem and Sklar, Stanford calls for a New Induction from the history of science. It will provide proof, he thinks, of “the kind of underdetermination that the history of science reveals to be a distinctive and genuine threat to even our best scientific theories” [Sta01, p. S12]. This paper examines Stanford’s New Induction and argues that it— like the other forms of underdetermination that he criticizes— merely recapitulates familiar philosophical conundra.

OL PCUCL PCP!0LT flJQOL!Q MJJICp MocTq Q2I4C qJQ uccq OL QUQL qw bcjj?ou bLoqnci bqcc opLAf!ou JAQj uq lUUGL I!c oj qW cxccLp!uWCLGPX oqici.bLopJcw 9uq bGc!Ucouv j qJ9I L c9Lc1Jc12 !U p.?4u o GAQ qJC urnJwuc bLopJcw jrni pJcq o wc n o 1p!uUGL UJTCLO qiw 9L bJ7UI2 01.JUJJa uq OWC WOLG LGCCIJI wni-cdnou cxicuou oj, w iuq DOflI iioqc wq ipc LcbouG o r /I JOoJ u bi.j1cnj qic tAOLJC qrn n bucj qw ou 0Cf1U buurnuj?OIJ jJJC JflQ 0j !UJnJWIJGA miq LCACP14U WQ 2LQUJ O, C14CIJJ O, 20W o qi couowcp.cpwou o btoqnciou flUC10U2 qicnnq VB2LKVCI 2EVKCH EOB IDEIAJII4CVJJOM bBODflGLIOM EflMCIIOI42: JJj NLCP I'IBEI f04C1U b'2OQ qtrj uq o c frrou9J ciciicc 9uq j0ffiJ 0flU410112 OL Q1191JC19J 21thb014 nuqj uqr J9jCG2 OL pcjbWj comujcu oujA cpou OMp1Cp conjq pG 1ucothoLGq nr JJT2 VAh klGAO t.O PJ IG2GC 92221JCG 10 0 CI J91U if CL)011 if H H1JL I K1U K H'-'qfl1JYGL2!IA iq 14BE1 3 uq CLG2 LI42EE b9L12 uq J'IBEW jc c rnqcpeq 1o bicrrjt q 2G1 CLO22 q1GIGU1 QUIJ2 uq q1cLcui 1]UJG bGLI0q2 DOG2 W anmcq jjJ9i 2 u MC 9W bbLOx1urnpu MGU MG NUC qowUJ O=EQC r ) uq bbjA i umupct oqu,cJij couccbnsj uq q bLopJcIu2: (j)12MJ MG 9W joojau oL Lc9JJA orr iu we cmbiucj mibjcuicunou 0LCCOU0UJGL!C bLoqncpou IPUCIIOU2 ouc 1mJ iuo cAcu rqeuppjc MIW w qu jrnuq dcrc2po1nu ocncq bum9uJA 011 pciyct we LcJf1oIi2pth OLflUCLC2 12 LcflX 1TqcujJcq, OL 10LCU20u (Jo8Q) 'ox.g LcMCM) if op sbbLocpc2 pvc nuqci.ouc2TuTcsU ccououJcpJc pucfrousJ LUJ2 qcAcJobeq lu bswjjc pn MC ij oujA oITcp ou i iu biu GW (2cc 111 W" G2WJJ1OU q 1qij11 JTfC1S1tC 011 C02 IfflJCT01J2 WCOL qcuuq 2A2GI1J2 ffIIq jexrpjc UCT0U2 pJCJ 92 mc92m.cWG1JLLI1JGM0LJ 0L btoqncp,A 9iJq jcq 1o S LC2111CUCC 0UUGLG2f sbb;icq O pe WOW bJSn2ipJc jpc LO1U 1UGW2 iu ccpuTcSJ cp9uc pLorrp bioqncrou (io) cxsuibjc MJJCLG we UGcG22SLX Bnmubpou2 OL c2purnpu 20WG1PTU cuipjc ILJICLO c2bCCIffjjA w 9&TCrrJfraG (2cc jjupct (JtzY nuqjsjc (joj) suq HCSqA suq jyjjou qr VIJGL S IJnuJpcL MujJCUU sgscjc pA CUIIC2 [J!2 IUJG OM0LJ( 2pIJeq LOW WSCLo o M0LJ0U2 0jIuffL1119j bLoqncpIuA WG0IA suq we coubcipAcuc92 OLISPOL myc fl21U D0ITTS2 (copp suq DO11J J8) jpcA as1.1cq oi1 s s ooj LOL c2Tu pThoWc2c2 spoiu we

After presenting the history, the evolution and the content of innovation surveys, we discuss the characteristics of the data they contain and the challenge they pose to the analyst and the econometrician.We document the two uses that have been made of these data: the construction of scoreboards for monitoring innovation and the scholarly analysis of various issue related to innovation.In particular we review the questions examined and the results obtained regarding the determinants, the effects, the complementarities, and the dynamics of innovation.We conclude by suggesting ways to improve the data collection and their econometric analysis.

In the Netherlands, the average exit rate out of welfare is dramatically low. Most welfare recipients have to comply with guidelines on job search effort that are imposed by the welfare agency. If they do not, then a sanction in the form of a temporary benefit reduction can be imposed. This article investigates the effect of such sanctions on the transition rate from welfare to work using a unique set of rich register data on welfare recipients. We find that the imposition of sanctions substantially increases the individual transition rate from welfare to work.

This article investigates estimation of finite population totals in the presence of univariate or multivariate auxiliary information. Estimation is equivalent to attaching weights to the survey data. We focus attention on the several weighting systems that can be associated with a given amount of auxiliary information and derive a weighting system with the aid of a distance measure and a set of calibration equations. We briefly mention an application to the case in which the information consists of known marginal counts in a two- or multi-way table, known as generalized raking. The general regression estimator (GREG) was conceived with multivariate auxiliary information in mind. Ordinarily, this estimator is justified by a regression relationship between the study variable y and the auxiliary vector x. But we note that the GREG can be derived by a different route by focusing instead on the weights. The ordinary sampling weights of the kth observation is 1/πk , where πk is the inclusion probability of k. We show that the weights implied by the GREG are as close as possible, according to a given distance measure, to the 1/πk while respecting side conditions called calibration equations. These state that the sample sum of the weighted auxiliary variable values must equal the known population total for that auxiliary variable. That is, the calibrated weights must give perfect estimates when applied to each auxiliary variable. That is a consistency check that appeals to many practitioners, because a strong correlation between the auxiliary variables and the study variable means that the weights that perform well for the auxiliary variable also should perform well for the study variable. The GREG uses the auxiliary information efficiently, so the estimates are precise; however, the individual weights are not always without reproach. For example, negative weights can occur, and in some applications this does not make sense. It is natural to seek the root of the dissatisfaction in the underlying distance measure. Consequently, we allow alternative distance measures that satisfy only a set of minimal requirements. Each distance measure leads, via the calibration equations, to a specific weighting system and thereby to a new estimator. These estimators form a family of calibration estimators. We show that the GREG is a first approximation to all other members of the family; all are asymptotically equivalent to the GREG, and the variance estimator already known for the GREG is recommended for use in any other member of the family. Numerical features of the weights and ease of computation become more than anything else the bases for choosing between the estimators. The reasoning is applied to calibration on known marginals of a two-way frequency table. Our family of distance measures leads in this case to a family of generalized raking procedures, of which classical raking ratio is one.