Search

We take a fresh look at Theil's BLUS residuals and ask why they have gone out of fashion.All our simulation experiments indicate that tests based on BLUS residuals have higher power than those based on the more popular recursive residuals, even in those cases (structural breaks) where intuition would favour the recursive residuals.

This article uses the Case-Shiller technique for constructing housing price indices on a Norwegian data set of transactions for the period 1991-2002 consisting of 10 376 pairs of repeated sales. Using a weighted least squares scheme in order to control for heteroskedasticity, we construct a general housing price index by regressing differences in log prices for the subset of repeated sales of same, and thus identical, homes onto a set of binary time variables, one for each quarter in the period. The constructed index shows that nominal prices for identical homes in general have increased by a factor of 3.58 over the 11-year period, while the CPI increased by 1.28, creating substantial capital returns for early purchasers. We then segment the data set into five different housing types in order to control for finite mixtures of hedonic features, and find that price indices for the smallest and largest type show nominal increases by factors 4.40 and 2.77, respectively. Keywords: distribution, hedonic model, housing price bubble, housing price index, inequality, repeated sales model, segmented housing types

The analysis of the exchange rate in Latvia presented in this publication covers the period from 1994 until 2001. The analysis has been based on two commonly used methods: the single equation approach and macroeconomic balance approach. The existing fundamental variables in the economy warrant the exchange rate that is undervalued relative to its equilibrium. Therefore, the current exchange rate in Latvia generally corresponds to the current stage of economic development. The real appreciation of the exchange rate with respect to West European countries is in line with the appreciation of the trend exchange rate, which is driven mainly by rising productivity in the tradable sector. Therefore, the appreciation of the real exchange rate from 1994 to 2001 has not generally harmed foreign trade. As long as real appreciation is supported by underlying fundamental variables in the economy, it does not endanger the macroeconomic stability and growth.

This paper questions traditional approaches for testing the day-of-the-week effect on stock returns. We propose an alternative approach based on the closure test principle introduced by Marcus, Peritz and Gabriel (1976), which has become very popular in Biometrics and Medical Statistics. We test all pairwise comparisons of daily expected stock returns, while the probability of committing any type I error is always kept smaller than or equal to some prespecified level a for each combination of true null hypotheses. We confirm day-of-theweek effects for the S&P 500, the FTSE 30 and the DAX 30 found in earlier studies, but find no evidence for the 1990's.

Various methods were proposed to understand the linkages in an input- output system; however many focused only on the identification of key sectors in the economy. An alternative approach, identifying analytically importance of elements and combinations of elements was proposed as a field of influence theory (Sonis et al., 1996). The purpose of this paper is to offer a complementary approach to the field of influence and the socalled 'Matrioshka principal' (Sonis and Hewings, 1990); the objectives are to identify simple row-column associations (i.e. statistical dependence), seek hierarchical associations between supply and demand in input-output systems and the decomposition of economic complexity into finite stages. For the identification of simple dependencies between rows and columns, we use a log-linear regression and for hierarchical associations and the identification of complexity stages, we use the data analysis technique known as dual scaling. Results of both approaches will be applied to input-output tables of the US and Canada.

Beginning with the recent reorganisation of the principal Italian public research bodies, which has led to the founding of larger labs, this paper evaluates the presence of statistically significant relationships between the dimensions of the research bodies and the performance of the same. The results of this research can be divided into three parts. The first concerns the descriptive-explorative analysis of the data; the second and third sections, based on the aforementioned data, will present the principal results obtained from the modelling activities with multiple regression and inference, both aimed at analysing the dynamics of the performance as the dimensions of the organisation change.

This paper analyses the impact of public innovation subsidies on private innovation expenditure. In the empirical economic literature there is still no common support for the hypothesis of either a complementary or a substitutive relationship between public funding and private investment. We investigate whether firms of the German service sector increase their innovation effort when participating in public policy schemes. Cross-sectional data at the firm level are used to estimate the effect of subsidization. Applying a non-parametric matching approach we find evidence that the hypothesis of complete crowding-out effects between public and private funds can be rejected.

This paper tries to distinguish between age and generation effects in the life cycle evolution of living standards of individuals. It is based on biographical data from a survey of 1998 made in Antananarivo and provided by the Madio project in Madagascar. Between 1965 and 1995, we observe both a diminution of living standards and an increase of inequalities in Antananarivo. In 35 years, the decrease of consumption has been continuous: the real per capita consumption of households decreased by 44.5 % between 1961 and 1995. At the same time, per capita GDP decreased by 36.8 % at the national level and private consumption by 46.8 %. In this context, what was the evolution of life cycle living standards of individuals? Have the youngest lower income compared to the elder at the same age? The analysis consists in the elaboration of a method of estimation of incomes during life trajectories of individuals of three different and successive generations. A double approach is used: a monetary analysis and another which takes into account living conditions of surveyed individuals during all their life. The evolution of inequality during the 1968-1998 period based on the preceding method is also studied to answer to two fundamental questions: which mobility in income scale do we observe in a long time period? What are the inequalities between individuals and cohorts when we evaluate it on the all life cycle? Finally, in the recent period (1988-1998), we shed the light on the education differentials between cohorts in the study of inequalities. _________________________________ Cette analyse tente de distinguer les effets d’âge et de génération dans l’évolution des niveaux de vie des individus sur le cycle de vie grâce à l’exploitation d’une enquête biographique réalisée en 1998 dans l’agglomération d’Antananarivo dans le cadre du projet Madio. Sur la période 1960-1995, on observe à la fois une chute des niveaux de vie et un accroissement des inégalités à Antananarivo. En 35 ans, la régression de la consommation a été continue: la consommation par tête des ménages, en volume, a chuté de 44,5 % entre 1961 et 1995. Dans le même temps, au niveau national, le PIB par tête a chuté de 36,8 % et la consommation privée de 46,8 %. Dans ce contexte, comment le niveau de vie a-t-il varié au cours du cycle de vie des individus? Les jeunes plus nombreux perçoivent–ils des revenus inférieurs à ceux de leurs aînés au même âge? L’analyse menée consiste en l’élaboration d’une méthode d’estimation des revenus sur les trajectoires de vie d’individus appartenant à trois générations successives. Une double approche est retenue: une analyse monétaire et une prise en compte des conditions de vie des individus enquêtés. L’évolution des inégalités sur la période 1968- 1998 sur la base de la méthodologie retenue est également étudiée pour répondre à deux questions essentielles : quelle mobilité dans l’échelle des revenus observe-t-on sur longue période ? Quelles sont les inégalités entre les individus ou les cohortes lorsqu’on les évalue sur tout le cycle de vie ? Enfin, la période récente (1988-1998) fait l’objet d’un éclairage particulier sur l’éducation différenciée des cohortes dans l’étude des inégalités.

This paper studies the Ordinary Least-Squared (OLS) and Instrumental Variables (IV) estimates of the returns to schooling for male workers in Spain. OLS estimates are often biased due to the endogeneity of schooling, measurement errors or omitted variables. Proper IV estimates correct this bias. The reliability of family background, natural experiments (based on changes in the education system and season of birth) and the availability of a college in the province is checked using Spanish data. The results suggest that background and college availability are valid instruments and that the IV estimates of the returns to schooling are higher than OLS estimates. These results are in line with the majority of previous results in the literature.

Several estimators have been suggested to tackle the problem of endogenous regressors and selectivity in count regression models. They differ in the structure and the degree of parametrization of the underlying models. The estimation of health services utilization conditional on the choice of different forms of health insurance provides a classical example of such problems. In Switzerland, basic health insurance is mandatory and each individual is insured separately. The insurance premium varies by region of residence but is independent of income and risk. The insured face a minimal annual deductible for ambulatory health services. Annually, they are given a choice of higher deductibles to reduce their insurance premium by a regulated percentage. The choice of a higher deductible sets incentives for a more cautious utilization of health services. Clearly, the choice is made based on expected health service utilization. The effect of the choice of a higher than the minimal deductible on the number of physician visits is analyzed. A matching estimator, a GMM estimator, two-stage method of moments estimators which account for selectivity and endogenous switching count regression models are applied to data from the 1997 Swiss Health Survey. Incentive-induced behavioral changes are disentangled from selection effects. The main finding is that most of the observed lower utilization for individuals with a high insurance deductible is caused by self- selection of individuals into the respective insurance contracts which either differ in their preferences or are healthier in unobserved aspects of their health status.

This paper presents a model of vehicle choice and then empirically examines the risk posed by light trucks (sport-utility vehicles, vans, and pickups) to those that drive them and to other drivers, relative to the risk posed by cars. The paper examines both the relative risk of dying given a crash as well as the relative crash frequencies of light trucks versus cars. The identification strategy uses information on pedestrian fatalities by vehicle type to correct for the sample selection bias that may exist due to the lack of reliable data on non-fatal crashes. Using data on all two-vehicle fatal crashes from 1991 through 1998, the results suggest that, given a crash, a light truck driver is 0.29 to 0.69 times as likely to die than is a car driver. On the other hand, given a crash, a light truck driver is 1.48 to 2.63 times as likely to kill the opposing driver than is a car driver. Using data from 1991 through 1994, the crash frequency estimates suggest that light trucks are approximately 2.2 times as likely to get into a crash than are cars. The relative safety of utility vehicles and pickups (compared to cars) disappears once one factors in the greater crash frequencies of light trucks. Factoring in the greater crash frequency of light trucks also increases their relative external risk. Light trucks are 3.26 to 5.78 times as likely to kill another driver than are cars. A world in which everyone drove light trucks would result in 2.81 to 6.31 times as many fatalities than a world in which everyone drove cars.

This paper provides empirical evidence on life-cycle patterns in the asset allocation of Swedish households. Data on household portfolio allocation are collected from the HINK surveys for the period 1982-1992, and portfolio shares of different asset categories are regressed on age, period, and cohort dummies as well as socio-economic and demographic variables. There are evident differences in the age profiles for the demand of different assets. The fraction of "risky" financial assets follows a hump-shaped age profile, as does the share of total real assets. While the probability of ownership of "safe" financial assets increases over life, the weight in the portfolio has a U-shaped age pattern. This is also true for the fraction of total financial assets. Furthermore, there are differences in the asset allocation of different birth cohorts; the portfolio weight of real assets is relatively higher for the "baby-boom" generation, while younger generations are more prone to invest in "risky" financial assets.

This paper examines conditional convergence of OECD countries in gross domestic product (GDP) and health care expenditure (HCE) per capita. It presents estimation of the augmented Solow growth model suggested by Mankiw, Romer and Weil (1992) to explain variation in output and expenditure per capita across countries. The variation is due to different steady state growth paths resulting from differences in the countries savings rate, education, and population growth. This paper is an extension of the MRW model by incorporating health capital proxied by HCE to the augmented Solow model. The analysis is further related to the studies of health care expenditure where GDP per capita appear to be the main factor determining the level of expenditure on health care. The issue of causality relationship between GDP and HCE is investigated. The empirical analysis is based OECD countries’ data for the period of 1970-1992. The results indicate that OECD countries converge at 3.7% per year to their steady state level of income per capita. The results show that HCE has positive effect on the economic growth and the speed of convergence. The speed of convergence is found to be sensitive to whether one imposes a constant or estimate the depreciation and technological growth components. With no restrictions imposed the convergence rate is 5.2%. Considering the rate of convergence in the HCE model the results show that OECD countries converge at 2.7% to their steady state of HCE per capita. In the HCE model a regression of the speed of convergence on variables determining the rate of convergence show close link to the variables characterizing the health care system of sample countries.

The least squares estimator is probably the most frequently used estimation method in regression analysis. Unfortunately, it is also quite sensitive to data contamination and model misspecification. Although there are several robust estimators designed for parametric regression models that can be used in place of least squares, these robust estimators cannot be easily applied to models containing binary and categorical explanatory variables. Therefore, I design a robust estimator that can be used for any linear regression model no matter what kind of explanatory variables the model contains. Additionally, I propose an adaptive procedure that maximizes the efficiency of the proposed estimator for a given data set while preserving its robustness.

This paper re-examines the issue of asymmetries in the transmission of shocks to crude oil prices onto the retail price of gasoline. Relative to the previous literature, the distinguishing features of the present paper are - i) use of updated and comparable data to carry out an international comparison of gasoline markets; ii) two-stage modeling of the transmission of oil price shocks to gasoline prices (first refinery stage and second distribution stage), in order to assess possible asymmetries at either one or both stages; iii) use of asymmetric error correction models to distinguish between asymmetries that arise from short-run deviations in input prices and from the speed at which the gasoline price reverts to its long-run level; iv) explicit, possibly asymmetric, role of the exchange rate, as crude oil is paid for in dollars whereas gasoline sells for different sums of national currencies; v) bootstrapping of F tests of asymmetries, in order to overcome the lowpower problem of conventional testing procedures. In contrast to several previous findings, the results generally point to widespread differences in both adjustment speeds and short-run responses when input prices rise or fall.

Dans ce texte, nous analysons les développements récents de l’économétrie à la lumière de la théorie des tests statistiques. Nous revoyons d’abord quelques principes fondamentaux de philosophie des sciences et de théorie statistique, en mettant l’accent sur la parcimonie et la falsifiabilité comme critères d’évaluation des modèles, sur le rôle de la théorie des tests comme formalisation du principe de falsification de modèles probabilistes, ainsi que sur la justification logique des notions de base de la théorie des tests (tel le niveau d’un test). Nous montrons ensuite que certaines des méthodes statistiques et économétriques les plus utilisées sont fondamentalement inappropriées pour les problèmes et modèles considérés, tandis que de nombreuses hypothèses, pour lesquelles des procédures de test sont communément proposées, ne sont en fait pas du tout testables. De telles situations conduisent à des problèmes statistiques mal posés. Nous analysons quelques cas particuliers de tels problèmes : (1) la construction d’intervalles de confiance dans le cadre de modèles structurels qui posent des problèmes d’identification; (2) la construction de tests pour des hypothèses non paramétriques, incluant la construction de procédures robustes à l’hétéroscédasticité, à la non-normalité ou à la spécification dynamique. Nous indiquons que ces difficultés proviennent souvent de l’ambition d’affaiblir les conditions de régularité nécessaires à toute analyse statistique ainsi que d’une utilisation inappropriée de résultats de théorie distributionnelle asymptotique. Enfin, nous soulignons l’importance de formuler des hypothèses et modèles testables, et de proposer des techniques économétriques dont les propriétés sont démontrables dans les échantillons finis., In this text, we review recent developments in econometrics from the viewpoint of statistical test theory. We first review some basic principles of philosophy of science and statistical theory, emphasizing parsimony and falsifiability as criteria for evaluating models, test theory as a formalization of the falsification principle for probabilistic models, and the logical foundation of basic notions in test theory (such as the level of a test). We then show that some of the most frequently used statistical and econometric methods are fundamentally inappropriate for the problems and models considered, while several hypotheses, for which test procedures are commonly proposed, are not testable at all. Such situations lead to ill-defined statistical problems. We analyze several cases of such problems : (1) building confidence intervals in structural models where identification problems may be present; (2) the construction of tests for nonparametric hypotheses, including procedures robust to heteroskadasticity, non-normality or dynamic specification. We point out that these difficulties often originate from the ambition to weaken regularity conditions typically required by statistical analysis, and from an inappropriate use of asymptotic distributional theory. Finally, we underscore the importance of formulating testable hypotheses and models, and of developing econometric methods with provable finite-sample properties. Keywords: econometrics; statistics; testing; philosophy of science; falsifiability; Popper; parsimony; identification; weak instruments; nonparametric model; time series; autoregressive model; heteroskedasticity; unit root; robust method; non-normality; ill-defined problem

In this article, we study the impact of payroll tax subsidies for low-wage workers on various outcomes of the firms, including employment, in particular that of young and less skilled workers. We concentrate on the effects of the 1995 and 1996 tax cuts policies, which permit large decreases in employer-paid contributions compared with the 1993 original policy. To distinguish between firms that are more or less concerned by these reductions, we compute for each firm in 1994 the changes in total labor costs, which are solely due to the changes in the tax reductions between 1994 and 1997. This variable lies between 0 and 10%, depending on the proportion of low wage workers in the firm in 1994. This paper includes an important methodology part. It refers to the statistical framework of Rubin (1974, 1977, 1983) and Heckman, Ichimura and Todd (1997, 1998, 1999), appropriate for the evaluation of a unique treatment, like participation in a training program. For identifying the effects of the firms ex ante labor cost reduction, we adapt the formalism to the case where the economic policies involve an infinite number of possible treatments. We also propose an estimation method based on the implementation of nonparametric series estimators. The empirical analysis makes use of matched employer-employee information originating from two main sources of data, the Déclarations Annuelles de Données Sociales (DADS) and the Bénéfices réels normaux (BRN). We find that, between 1994 and 1997, tax reductions are associated with very strong employment and wages effects in the economy. We also find that they permit the creation or the save of roughly 400.000 works, even though this estimate is rather imprecise because of the methodology.

This paper proves a new representation theorem for domains with both discrete and continuous variables. The result generalizes Debreu's well-known representation theorem on connected domains. A strengthening of the standard continuity axiom is used in order to guarantee the existence of a representation. A generalization of the main theorem and an application of the more general result are also presented., Ce papier prouve un nouveau théorème de représentation pour des domaines avec des variables tant discrètes que continues. Le résultat généralise le théorème de représentation bien connu de Debreu dans des domaines connexes. Un renforcement de l'axiome de continuité standard est employé pour garantir l'existence d'une représentation. Une généralisation du théorème principal et une application du résultat plus général sont aussi présentées.

We examine the relationship between the risk premium on the S&P 500 index return and its conditional variance. We use the SMEGARCH - Semiparametric-Mean EGARCH - model in which the conditional variance process is EGARCH while the conditional mean is an arbitrary function of the conditional variance. For monthly S&P 500 excess returns, the relationship between the two moments that we uncover is nonlinear and nonmonotonic. Moreover, we find considerable persistence in the conditional variance as well as a leverage effect, as documented by others. Moreover, the shape of these relationships seems to be relatively stable over time., Nous étudions la relation entre la prime de risque sur l'indice S&P 500 et sa variance conditionnelle. Nous utilisons le modèle SMEGARCH - Semiparametric-Mean EGARCH - selon lequel la variance conditionnelle suit un processus EGARCH, alors que la moyenne est une fonction arbitraire de la variance conditionnelle. Pour les rendements excédentaires mensuels sur l'indice S&P 500, la relation que nous trouvons est non linéaire et non monotone. De plus, nous trouvons beaucoup de persistance dans la variance conditionnelle ainsi qu'un effet de levier, tel que documenté par plusieurs autres auteurs.

This paper documents the high level of job security protection in Latin American labor markets and analyzes its impacts on employment. The authors show that job security policies have a substantial impact on the level and the distribution of employment in Latin America. These policies reduce employment and promote inequality. The institutional organization of the labor market affects both employment and inequality.

The purpose of this study was to investigate the contributions of Social Identify Theory’s variables to research performance. Regression models were applied on data of ten national research Council (Cnr) institutes of Piemonte, a highly industrialised region in North Western Italy.Results show that there is a low but positive correlation between performance and work satisfaction, which, the latter, has a strong positive correlation with the work involvement. Implications for R&D management are discussed

In Potscher (1991, Econometric Theory 7, 163–185) the asymptotic distribution of a post-model-selection estimator, both unconditional and conditional on selecting a correct model (minimal or not), has been derived. Limitations of these results are (i) that they do not provide information on the distribution of the post-model-selection estimator conditional on selecting an incorrect model and (ii) that the quality of this asymptotic approximation to the finite-sample distribution is not uniform with respect to the underlying parameters. In the present paper we first obtain the unconditional and also the conditional finite-sample distribution of the post-model-selection estimator, which turn out to be complicated and difficult to interpret. Second, we obtain approximations to the finite-sample distributions that are as simple and easy to interpret as the asymptotic distributions obtained in Potscher (1991) but at the same time are close to the finite-sample distributions uniformly with respect to the underlying parameters. As a by-product, we also obtain the asymptotic distribution conditional on selecting an incorrect model.We thank the co-editor Richard Smith and the two referees for helpful comments on a previous version of this paper. Hannes Leeb's research was supported by the Austrian Science Foundation (FWF), project P13868-MAT.

During the last 20 years, R&D and innovation activities in the service sector have clearly increased. Especially business services are believed to be one of the main drivers of technical changes and economic progress. Looking at the labour indices calculated over the period from 1982 to 1996 one notices a remarkable increase of over 70 percent for the business services. About 8 percent of total employment in West Germany is in business services. In particular, by taking advantage of information and communication technologies, knowledge-intensive business service firms increasingly play the role of "converters" of technological information within the economy. They are providers, purchasers or partners in the context of innovation. A sound innovation capacity, especially knowledge, creativity, market and management skills let them become bridges for innovation.

In the context of multivariate linear regression (MLR) models, it is well known that commonly employed asymptotic test criteria are seriously biased towards overrejection. In this paper, we propose a general method for constructing exact tests of possibly nonlinear hypotheses on the coefficients of MLR systems. For the case of uniform linear hypotheses, we present exact distributional invariance results concerning several standard test criteria. These include Wilks' likelihood ratio (LR) criterion as well as trace and maximum root criteria. The normality assumption is not necessary for most of the results to hold. Implications for inference are two-fold. First, invariance to nuisance parameters entails that the technique of Monte Carlo tests can be applied on all these statistics to obtain exact tests of uniform linear hypotheses. Second, the invariance property of the latter statistic is exploited to derive general nuisance-parameter-free bounds on the distribution of the LR statistic for arbitrary hypotheses. Even though it may be difficult to compute these bounds analytically, they can easily be simulated, hence yielding exact bounds Monte Carlo tests. Illustrative simulation experiments show that the bounds are sufficiently tight to provide conclusive results with a high probability. Our findings illustrate the value of the bounds as a tool to be used in conjunction with more traditional simulation-based test methods (e.g., the parametric bootstrap) which may be applied when the bounds are not conclusive., Dans le contexte des modèles de régression multivariés (MLR), il est bien connu que les tests asymptotiques usuels tendent à rejeter trop souvent les hypothèses considérées. Dans cet article, nous proposons une méthode générale qui permet de construire des tests exacts pour des hypothèses possiblement non linéaires sur les coefficients de tels modèles. Pour le cas des hypothèses uniformes linéaires, nous présentons des résultats sur la distribution exacte de plusieurs statistiques de test usuelles. Ces dernières incluent le critère du quotient de vraisemblance (Wilks), de même que les critères de la trace et de la racine maximale. L'hypothèse de normalité des erreurs n'est pas requise pour la plupart des résultats présentés. Ceux-ci ont deux types de conséquences pour l'inférence statistique. Premièrement, l'invariance par rapport aux paramètres de nuisance signifie que l'on peut appliquer la technique des tests de Monte Carlo afin de construire des tests exacts pour les hypothèses uniformes linéaires. Deuxièmement, nous montrons comment exploiter cette propriété afin d'obtenir des bornes sans paramètres de nuisance sur la distribution des statistiques de quotient de vraisemblance pour des hypothèses générales. Même si les bornes ne sont pas faciles à calculer par des moyens analytiques, on peut les simuler aisément et ainsi effectuer des tests de Monte Carlo à bornes. Nous présentons une expérience de simulation qui montre que ces bornes sont suffisamment serrées pour fournir des résultats concluants avec une forte probabilité. Nos résultats démontre la valeur de ces bornes comme instrument à utiliser conjointement avec des méthodes d'inférence simulée plus traditionnelles (telles que le bootstrap paramétrique) que l'on peut appliquer lorsque le test à bornes n'est pas concluant.

In this paper, we develop finite-sample inference procedures for stationary and nonstationary autoregressive (AR) models. The method is based on special properties of Markov processes and a split-sample technique. The results on Markovian processes (intercalary independence and truncation) only require the existence of conditional densities. They are proved for possibly nonstationary and/or non-Gaussian multivariate Markov processes. In the context of a linear regression model with AR(1) errors, we show how these results can be used to simplify the distributional properties of the model by conditioning a subset of the data on the remaining observations. This transformation leads to a new model which has the form of a two-sided autoregression to which standard classical linear regression inference techniques can be applied. We show how to derive tests and confidence sets for the mean and/or autoregressive parameters of the model. We also develop a test on the order of an autoregression. We show that a combination of subsample-based inferences can improve the performance of the procedure. An application to U.S. domestic investment data illustrates the method., Dans cet article, nous proposons des procédures d’inférence valides à distance finie pour des modèles autorégressifs (AR) stationnaires et non stationnaires. La méthode suggérée est fondée sur des propriétés particulières des processus markoviens combinées à une technique de subdivision d’échantillon. Les résultats sur les processus de Markov (indépendance intercalaire, troncature) ne requièrent que l’existence de densités conditionnelles. Nous démontrons les propriétés requises pour des processus markoviens multivariés possiblement non stationnaires et non gaussiens. Pour le cas des modèles de régression linéaires avec erreurs autorégressives d’ordre un, nous montrons comment utiliser ces résultats afin de simplifier les propriétés distributionnelles du modèle en considérant la distribution conditionnelle d’une partie des observations, étant donné le reste. Cette transformation conduit à un nouveau modèle qui a la forme d’une autorégression bilatérale à laquelle on peut appliquer les techniques usuelles d’analyse des modèles de régression linéaires. Nous montrons comment obtenir des tests et régions de confiance pour la moyenne et les paramètres autorégressifs du modèle. Nous proposons aussi un test pour l’ordre d’une autorégression. Nous montrons qu’une technique de combinaison de tests obtenus à partir de plusieurs sous-échantillons peut améliorer la performance de la procédure. Enfin, la méthode est appliquée à un modèle de l’investissement aux États-Unis.

Aim of this paper is to build a method of performance evaluation for research organization, using a systemic approach that considers the interrelations among activities (administrative, scientific, technologic). The method is based on the research laboratories evaluation (relev) function which is a linear combination of seven indices, two financials, two tacit technological transfer, two bibliometrics and one technological. The relev function summarises quantitative, qualitative and cost aspects, it is simple and emphasizes evaluation of R&D outcomes rather than behaviour. The validity of the relev methodology was tested on the National Research Council Institutes operating in North West Italy.

In econometric analysis, non-nested models arise naturally when rival economic theories are used to explain the same phenomenon, such as unemployment, inflation or output growth. The authors examine the problem of hypothesis testing when the models under consideration are ‘non-nested’ or belong to ‘separate’ families of distributions in the sense that none of the individual models may be obtained form the remaining, either by imposition of parameter restrictions or through a limiting process. Although the primary focus is on non-nested hypothesis testing, the authors briefly discuss the problem of model selection and the differences and similarities between the two approaches. By using the linear regression model as a convenient framework, the authors examine three broad approaches to non-nested hypothesis testing: the modified (centred) long-likelihood ratio procedure, the comprehensive models approach, and the encompassing procedure. Finally, they consider a number of practical problems which arise in the application of non-nested tests to non-linear models such as the probit and logit qualitative response models.

In this paper we construct a test for the difference parameter d in the fractionally integrated autoregressive moving-average (ARFIMA) model. Obtaining estimates by smoothed spectral regression estimation method, we use the moving blocks bootstrap method to construct the test for d. The results of Monte Carlo studies show that this test is generally valid for certain block sizes, and for these block sizes, the test has reasonably good power.

Measurement of seat belt and air bag effectiveness is complicated by the fact that systematic data are collected only for crashes in which a fatality occurs. These data suffer from sample selection since seat belt and air bag usage influences survival rates which in turn determine whether a crash is included in the sample. Past researchers either ignored sample selection or adopted indirect estimation methods subject to other important biases. We propose a simple, but novel, solution to the selection problem: limiting the sample to crashes in which someone in a different vehicle dies. Under relatively weak conditions, consistent estimates can be obtained from this restricted sample. Empirically, we find seat belts to be more effective in saving lives than most previous estimates. Air bags, however, appear to be less effective than generally thought. If our coefficients can be generalized to all crashes, the cost per life saved with seat belts is approximately $30,000, compared to $1.6 million for air bags.

The paper proposes an instrumental variables version of the Huber estimator as an alternative to the IV-Krasker Welsch estimator. The IV-Huber estimator is analytically and computationally much simpler than IV-Krasker Welsch. In the context of an empirical study of the importance of borrowing constraints on consumption, the paper reports the results for the following estimators: 1) conventional (non-robust) IV, 2) conventional IV after the subjective rejection of outliers, 3) conventional IV after trimming, 4) IV-Huber, and 5) IV-Krasker-Welsch. In the presence of a heavy-tailed error distribution, both the IV-Krasker Welsch and the IV-Huber estimators provide substantial improvements in efficiency over conventional IV. Further, the informal robust procedure of using conventional IV after trimming does not match the efficiency gains of the formal robust methods. The empirical results indicate that households exhibit incomplete smoothing of consumption, with about 20-50% of predictable movements in income being buffered by asset stocks. When saving is disaggregated by type of asset, the results provide some evidence of borrowing constraints: households which are not subject to a liquidity constraint use financial assets as their primary means of buffering income fluctuations, while constrained households use purchases of durable goods almost exclusively as the vehicle for consumption smoothing.

Recent work suggests that the conditional variance of financial returns may exhibit sudden jumps. This paper extends a non-parametric procedure to detect discontinuities in otherwise continuous functions of a random variable developed by Delgado and Hidalgo (1996) to higher conditional moments, in particular the conditional variance. Simulation results show that the procedure provides reasonable estimates of the number and location of jumps. This procedure detects several jumps in the conditional variance of daily returns on the S&P 500 index., Des études récentes suggèrent que la variance conditionnelle des rendements financiers est sujette à des sauts. Ce papier étend une procédure non paramétrique de détection de sauts développée par Delgado et Hidalgo (1996) à la détection de sauts dans les moments conditionnels supérieurs, en particulier la variance conditionnelle. Les résultats de simulation démontrent que cette procédure estime de façon raisonnable le nombre de sauts ainsi que leurs emplacements. L'application de cette procédure aux rendements journaliers sur l'indice S&P 500 révèle la présence de plusieurs sauts dans la variance conditionnelle.

In this paper, the role of the computer at the workplace will be examined in determining the wage structure in Germany. Following Krueger (1993) and using the German Socio-Economic Panel (GSOEP), cross-sectional wage regression results from 1997 and panel results from 1984-1997 are presented. It is shown that the wage premium attributed to using a computer at work using cross-sectional results for 1997 is around 7%. Further it is shown that computer usage is very heterogeneous depending on which industry one works in. In cross-section, hypothesis tests show that several industries and almost all firm size categories exhibit very different wage differentials depending on computer usage at the workplace. As DiNardo and Pischke (1997) stress the need for panel data to control for unmeasured individual effects, we use GSOEP 1984-1997 panel data, where a random effects and fixed effects estimator were run in the wage estimation. We confirm the results that Entorf and Kramarz (1997) had for France, that in Germany the coefficient for computer usage at the workplace did not remain stable and although just barely significant, was reduced to mere 1% with individual fixed effects. We conclude that there are no computer usage wage differentials worth speaking of, once one controls adequately for unobserved individual heterogeneity.

This paper condiders an extension of Tran Van Hoa's family of 2SHI (two stage hierarchical information) estimators for the coefficient vector of a linear regression model and derives the conditions for the dominance of 2SHI estimator over the OLS and Stein rule estimators under a Generalised Pitman Nearness (GPN) criterion when the disturbance variable is small.

Lam and Schoeni (1993) consider an equation where earnings are explained by schooling and ability. They assume that ability data are lacking and that schooling is measured with error. The estimate obtained by regressing earnings on schooling thus contains omitted variable bias (OVB), which is positive, and measurement error bias (MEB), which is negative. Adding a family background variable to proxy for ability is claimed to: i) decrease the OVB towards, but not below, zero and ii) make the MEB even more negative. This note claims that while ii) is true, even in the context of multiple family background variables, i) is in general incorrect. The OVB may increase in magnitude and/or change sign. Conditions are provided under which i) holds. A simulation procedure is suggested that will yield consistent estimates of the total bias and its components, conditional upon values on the true return and the measurement error variance.

In the context of multivariate regression (MLR) and seemingly unrelated regressions (SURE) models, it is well known that commonly employed asymptotic test criteria are seriously biased towards overrejection. in this paper, we propose finite-and large-sample likelihood-based test procedures for possibly non-linear hypotheses on the coefficients of MLR and SURE systems.

In a recent paper, Bai and Perron (1998) considered theoretical issues related to the limiting distribution of estimators and test statistics in the linear model with multiple structural changes. In this companion paper, we consider practical issues for the empirical applications of the procedures. We first address the problem of estimation of the break dates and present an efficient algorithm to obtain global minimizers of the sum of squared residuals. This algorithm is based on the principle of dynamic programming and requires at most least-squares operations of order O(T 2) for any number of breaks. Our method can be applied to both pure and partial structural-change models. Secondly, we consider the problem of forming confidence intervals for the break dates under various hypotheses about the structure of the data and the errors across segments. Third, we address the issue of testing for structural changes under very general conditions on the data and the errors. Fourth, we address the issue of estimating the number of breaks. We present simulation results pertaining to the behavior of the estimators and tests in finite samples. Finally, a few empirical applications are presented to illustrate the usefulness of the procedures. All methods discussed are implemented in a GAUSS program available upon request for non-profit academic use., Dans un récent papier, Bai et Perron (1998) ont considéré les problèmes théoriques reliés à la distribution des estimateurs et tests statistiques dans le modèle linéaire avec changements structurels multiples. Dans ce papier, nous regardons les problèmes pratiques pour les applications empiriques des procédures. En premier, nous regardons le problème d'estimation des dates de rupture et présentons un algorithme efficace pour obtenir les minimums globaux des sommes des résidus carrés. Cet algorithme est basé sur le principe de la programmation dynamique et nécessite au plus des opérations de moindres carrés d'ordre O(T 2) pour tout nombre de ruptures. Deuxièmement, nous considérons le problème de construire des intervalles de confiance pour les dates de rupture sous plusieurs hypothèses de structure des données et d'erreurs entre segments. Troisièmement, nous considérons le problème de tester la présence de changements structurels sous des conditions très générales sur les donnés et les erreurs. Quatrièmement, nous étudions l'estimation du nombre de ruptures. Nous présentons les résultats de simulations sur le comportement des estimateurs et des tests en échantillons finis. Finalement, nous offrons quelques applications empiriques pour illustrer l'utilité des procédures. Toutes les méthodes présentées sont exécutées à l'aide d'un programme GAUSS disponible sur demande pour utilisation académique seulement.

This note investigates the adequacy of the finite-sample approximation provided by the Functional Central Limit Theorem (FCLT) when the errors are allowed to be dependent. We compare the distribution of the scaled partial sums of some data with the distribution of the Wiener process to which it converges. Our setup is purposely very simple in that it considers data generated from an ARMA(1,1) process. Yet, this is sufficient to bring out interesting conclusions about the particular elements which cause the approximations to be inadequate in even quite large sample sizes., Cette note examine la pertinence de l'approximation en échantillon fini donnée par le théorème de la limite centrale fonctionnelle quand les erreurs sont dépendantes. Nous comparons la distribution des sommes partielles échelonnées de quelques donnés avec la distribution du processus de Wiener sur lequel elle converge. Notre modèle est intentionnellement très simple, compte tenu qu'il considère les donnés générées à partir d'un processus ARMA(1,1). Cependant, c'est suffisant pour obtenir des conclusions intéressantes sur les éléments particuliers qui causent des approximations inadéquates même en larges échantillons.

In the empirical literature on labour supply, several models are developed to incorporate constraints on working hours. These models do not address the question to which extent working hours are constrained within and between jobs. In this paper I investigate the effect of individual changes in labour supply preferences on actual working hours. The availability of subjective information on the individual’s preferred working hours gives direct measures on the degree of adjustment of working hours within and between jobs. I take the potential endogeneity of the observed changes in labour supply preferences into account by using instrumental variables techniques. Using the Dutch Socio Economic Panel, I find for women who work in two consecutive years that the degree of adjustment of working hours within jobs is significantly smaller than between jobs. I also find that job mobility does not lead to complete adjustment in working hours, as the realized adjustment is significantly less than 100 percent of the preferred adjustment.

This paper deals with the general partitioned linear regression model where the regressor matrix $X=\pmatrix{X_1 & X_2\cr}$ may be deficient in column rank, the dispersion matrix $V$ is possibly singular, $\beta^t=\pmatrix{\beta_1^t & \beta_2^t\cr}$ - being partitioned according to $X$ - is the vector of unknown regression coefficients, and $\beta_2$ is possibly subject to consistent linear equality or inequality restrictions. In particular, we are interested in the set of {\it generalized least squares (GLS) selections} for $\beta_2$. Inspired by Aigner and Balestra [1], as well as by Nurhonen and Puntanen [2], we also consider a specific reduced model and describe a scenario under which the set of GLS selections for $\beta_2$ under the reduced model equals the set of GLS selections for $\beta_2$ under the original full model. The results obtained in [2] and [1] for the unrestricted {\it standard} (full rank) regression model are reobtained as special cases.

The connection between fiscal policy and economic growth is not a strong one, takinginto account that fiscal policy is not a fundamental source of growth. Even so, governmentauthorities could use fiscal policy to affect in an indirect manner the economic growth. We will tryto highlight this indirect connection and its strength on Romania’s case using a specificeconometric methodology which takes into account the restrictions imposed by the government’sbudget constraint and we will identify the specific fiscal policy measures which could enhanceeconomic growth in Romania.

The link between investment decisions undertaken by economic agents and corporateincome (profit) taxation is well documented in theoretical studies realized so far. But, the empiricalevidence is very mixed and do not provide clear answers regarding the magnitude of the taxationeffects on investment and the proper transmission channels for these effects. So, we propose toinvestigate the effects of corporate income (profit) tax burden on investment decision for a samplecomposed from EU Member States.

No abstract.

This empirical study of Armington elasticities between Australian domestically produced commodities and imported commodities of the same name follows directly from Alaouze, Marsden and Zeitsch [AMZ] (1977) (see Impact Project Working Paper O-11, available from this web site). In the present study the AMZ models were estimated from data constructed at the input-output [IO] (rather than the four-digit ASIC) level of aggregation. The major results obtained here are in basic agreement with those in AMZ. However, estimated substitution elasticities at this higher level of aggregation tended to be lower than those obtained in AMZ. Nevertheless, the elasticities of substitution estimated here at the IO aggregation level fell within the range of the estimates for any given IO category's ASIC components in those cases where substitution elasticities for two or more ASICs belonging to the same IO category were estimated in the AMZ study.

This 1976 paper discusses a multi-input, multi-output production function for modelling Australian agriculture within the ORANI model. Literature is surveyed, and a theoretical approach outlined. A CRESH-CRETH production function is proposed, and equations derived. Theory and data are described, but actual regression results are deferred to a later report. Please note: The PDF download of this fairly old paper is an optical scan of indifferent, but legible, quality

This 1978 paper is about modelling production input demands for the ORANI model. It examines a nested production structure where capital and labour are combined to form "value-added", which is in turn combined with a materials aggregate to form output. The nested production structure is contrasted with a more general translog form. Using trans-log, own-and cross-price elasticities are estimated. The elasticity matrix does not show the patterns implied by the nesting hypothesis which is therefore rejected. Please note: The PDF download of this fairly old paper is an optical scan of indifferent, but legible, quality

This 1977 paper is one of several by Alaouze about estimating Armington elasticities for the ORANI model. It postulates that during periods of high demand, domestic producers are unable to supply goods promptly and so purchasers must wait for delivery. The implicit cost of waiting for domestic goods inclines users to imported equivalents, even if domestic prices do not rise relative to foreign. A measure or index for "pressure of demand" is described which is then used as a regressor to explain import demand. Please note: The PDF download of this fairly old paper is an optical scan of indifferent, but legible, quality

This 1978 paper is about estimating elasticities for the ORANI model -- in particular the elasticity of substitution between labour of different types. Literature is surveyed, and a theoretical approach outlined, but discussions of data and regression results are deferred to a later report. Please note: The PDF download of this fairly old paper is an optical scan of indifferent, but legible, quality