Success breeds success. The dynamics of the innovation process

This paper presents a dynamic random effects probit model for the realization of private firms' product and process innovations. We estimate the model with panel data collected by the Ifo Institute in Munich. The data covers the period between 1979 and 1986 and includes 301 firms of the West German manufacturing sector. It turns out that firms' probabilities of innovation depend on market structure, demand and cost expectations, unobserved heterogeneity, and realized innovations in the previous year. The positive significant influence of past innovations indicates that there is strong state dependence in the innovation process. This result supports the success breeds success hypothesis suggesting a positive impact of innovative success to further innovations in the following years.We would like to thank the Ifo-Institute for providing us with the data of the Ifo-Konjunkturtest. Special thanks to Georg Licht and Horst Rottmann for preparing the data. Helpful comments were received from participants of seminars in Mannheim, Hannover, and Nürnberg, especially from Dietmar Harhoff. We are particularly indebted to an anonymous referee for valuable suggestions. Financial support of the DFG is gratefully acknowledged.     

Inverse probability weighted M-estimators for sample selection, attrition, and stratification

I provide an overview of inverse probability weighted (IPW) M-estimators for cross section and two-period panel data applications. Under an ignorability assumption, I show that population parameters are identified, and provide straightforward -consistent and asymptotically normal estimation methods. I show that estimating a binary response selection model by conditional maximum likelihood leads to a more efficient estimator than using known probabilities, a result that unifies several disparate results in the literature. But IPW estimation is not a panacea: in some important cases of nonresponse, unweighted estimators will be consistent under weaker ignorability assumptions.JEL Classification: C13, C21, C23I would like to thank Bo Honoré, Christophe Muller, Frank Windmeijer, and the participants at the CeMMAP/ESCR Econometric Study Group Microeconometrics Workshop for helpful comments on an earlier draft.     

Nonparametric estimation of a dependent competing risks model for unemployment durations

In this paper we simultaneously analyze transitions from unemployment to employment and to nonparticipation. We estimate a dependent competing risks model with nonparametric specifications of the destination-specific duration dependence and unobserved heterogeneity terms, allowing for mutual dependence of the unobserved heterogeneity terms. We use an administrative data set covering all registered French unemployed over the period 1988–1994, stratified by gender type, duration class and exit state. We thank the Editor and three anonymous referees for their useful comments. A preliminary version of this paper was distributed under the title “Individual variation in exit rates from unemployment: a nonparametric multivariate analysis using aggregate data”. The Département de Marché du Travail of the Ministère du Travail, de l’Emploi et de la Formation Professionelle of France kindly provided the data.     

Semiparametric Bayesian estimation of mixed count regression models

This paper develops semiparametric Bayesian estimation approach for Poisson regression models with unobserved heterogeneity of unknown density. This approach is computationally efficient and allows automatic adaptation of the approximating density to data during estimation. Simulations show the estimator performs well.     

Subjective well-being and the family: Results from an ordered probit model with multiple random effects

The previous literature on the determinants of individual well-being has failed to fully account for the interdependencies in well-being at the family level. This paper develops an ordered probit model with multiple random effects that allows to identify the intra-family correlation in well-being. The parameters of the model can be estimated with panel data using Maximum Marginal Likelihood. The approach is illustrated in an application using data for the period 1984–1997 from the German Socio-Economic Panel in which both inter-generational and intra-marriage correlations in well-being are estimated.First version received: May 2003/Final version received: August 2004I would like to thank Andrew Clark, Daniel S. Hamermesh, seminar participants at Tilburg University, and three anonymous referees for helpful comments.     

Public capital stock and state productivity growth: Further evidence from an error components model

The contribution of different types of public infrastructure on private production is investigated using time-series of cross-section data for the 48 contiguous states over the period 1970–1986. A Cobb-Douglas production function is estimated with unobserved state-specific effects. Measurement errors in public capital stock and its components are detected and rectified.We would like to thank Baldev Raj and an anonymous referee for helpful comments. Also, Timothy J. Gronberg and Kay McAllister who thoroughly read the earlier draft and offered many constructive suggestions. We are, however, solely responsible for any remaining errors. The data set used in this research was generously provided by Alicia H. Munnell and Leah Cook of the Federal Reserve Bank of Boston. Nat Pinnoi acknowledges the research support provided by the Texas Transportation Institute.     

Efficiency measurement using a latent class stochastic frontier model

Efficiency estimation in stochastic frontier models typically assumes that the underlying production technology is the same for all firms. There might, however, be unobserved differences in technologies that might be inappropriately labeled as inefficiency if such variations in technology are not taken into account. We address this issue by estimating a latent class stochastic frontier model in a panel data framework. An application of the model is presented using Spanish banking data. Our results show that bank-heterogeneity can be fully controlled when a model with four classes is estimated. This paper was written during Luis Oreas visit to Binghamton University in the summer of 2002. We would like to thank an associate editor of the journal and two anonymous referees for their detailed comments. However, we alone are responsible for any remaining errors.     

Estimating the employment effects of education for disabled workers in Norway

In this paper we evaluate the effectiveness of educational programmes used as an employment strategy for disabled workers in Norway. To obtain these estimates we follow the employment career of a sample of participants in educational programmes and nonparticipants three years after they had left the vocational rehabilitation benefit system. We specify an employment outcome model that includes both unobserved heterogeneity and selection bias due to correlation between unmeasured factors and a person's training status. Even though participants in educational programmes have employment rates that are around eight percentage points higher than those who did not participate in such programmes, econometric selection models produce a training effect for education not significantly different from zero. First version: Jan 2001/Final version received: April 2002  The data were provided by The Norwegian Social Data Service in Bergen. I would like to thank Dag Kiberg for preparing the raw data. All the data are gathered from sources at The Directorate of Labour, The Social Insurance Organization and Statistics Norway. None of the mentioned institutions are in any circumstances responsible for the analysis or for the conclusions drawn from it.     

Sunk Costs,Firm Heterogeneity,Export Market Entry and Exit: Evidence from India

This paper analyses the role of sunk costs and firm heterogeneity in firm decision to enter and exit export markets. Employing rich firm-level data on Indian manufacturing firms, the study points out that sunk costs in terms of previous export experience significantly explain entry and exit decisions of firms in the export market. The first set of analysis involves estimation of dynamic discrete choice model using random effects probit correcting for initial conditions problem. We find evidence that previous export experience (sunk costs) matters for export decision. However, importance of sunk costs is found to depreciate rapidly. Further, analysis across sub-sample of firms accounting for firm heterogeneity factors like size and product level information supports the hypothesis of sunk costs. Second set of analysis involving firm survival in export markets using discrete-time hazard models shows evidence of negative duration dependence. We observe that those firms which continue to export for few years are less likely to exit from export markets.

     

WORK HISTORIES AND LIFETIME UNEMPLOYMENT

A long-standing question in economics is how important unobserved differences across workers are for explaining unemployment. I revisit this topic using variation in lifetime unemployment across workers in U.S. data. A comparison of workers often unemployed with the rest shows that although differences in job-finding rates increase over the course of a career, differences in job-separation rates are large right from the start. I develop a directed search model with symmetric unobserved heterogeneity, in which agents learn workers' types from their labor market histories, to rationalize these findings. The model cannot match the data if unobserved heterogeneity is neglected.     

The duration of trade revisited

The recent literature on the duration of trade has predominantly analyzed the determinants of trade flow durations using Cox proportional hazards models. The purpose of this article is to show why it is inappropriate to analyze the duration of trade with continuous-time models such as the Cox model, and to propose alternative discrete-time models which are more suitable for estimation. In brief, the Cox model has three major drawbacks when applied to large trade data sets. First, it faces problems in the presence of many tied duration times, leading to biased coefficient estimates and standard errors. Second, it is difficult to properly control for unobserved heterogeneity, which can lead to parameter bias and bias in the estimated survivor function. Third, the Cox model imposes the restrictive and empirically questionable assumption of proportional hazards. In contrast, with discrete-time models there is no problem handling ties; unobserved heterogeneity can be controlled for without difficulty; and the restrictive proportional hazards assumption can easily be bypassed. By replicating an influential study by Besede? and Prusa (J Int Econ 70:339?C358, 2006b), but employing discrete-time models as well as the original Cox model, we find empirical support for each of these arguments against the Cox model. Moreover, when comparing estimation results obtained from a Cox model and our preferred discrete-time specification, we find significant differences in both the predicted survivor functions and the estimated effects of explanatory variables on the hazard. In other words, the choice between models affects the economic conclusions that can be drawn.     

Alcohol content and demand for alcoholic beverages: A system approach

This paper examines how alcohol content affects the consumption of alcoholic beverages in Finland. Three different quality hypotheses are studied and compared: Fisher and Shell, Theil, and an additive one. The comparison of the hypotheses is based on quality elasticities implied by the hypotheses. The results show that, under all hypotheses, alcohol content positively affects the demand for alcoholic beverages, and this effect depends negatively on income. The results of the comparison of the hypotheses show that the additive fits the data best. However, the other hypotheses are almost as good: Fisher and Shell's hypothesis better than Theil's.I would like to thank K. Koskela, A. Nyberg, M. Salo, M. Stenius, and I. Suoniemi for their useful comments and suggestions. The author bears sole responsibility for any remaining errors.     

On the role of unobserved preference heterogeneity in discrete choice models of labour supply

The aim of this paper is to analyse the impact of unobserved preference heterogeneity in empirical applications of discrete choice models of labour supply. Typically, unobserved heterogeneity is estimated either with continuous or discrete mixture models. However, in order to avoid estimation difficulties, most of the empirical analysis assumes a relatively constrained mixture, standard examples being models where only few coefficients are allowed to vary with independent normal distributions or with discrete distributions with few mass points. We compare labour supply elasticities obtained with these typical specifications of unobserved heterogeneity with those from a more general model that we are able to estimate through an EM algorithm for the nonparametric estimation of mixed models. Results show that labour supply elasticities change significantly with respect to a basic model without unobserved heterogeneity only when the joint distribution of the varying tastes is left completely unspecified.     

Fertility,human capital and the wealth of families

Summary An overlapping generations model with parental altruism is examined. The existence of the optimal value function in a model with an endogenous discount rate is proven. Two development regimes are produced: a high fertility, low income and no growth steady state, and a perpetual growth equilibrium with low fertility and rising income.This paper is adapted from my dissertation. I would like to thank the members of my dissertation committee for helpful comments and suggestions, Messrs, Gary S. Becker, Robert E. Lucas, Jr., Kevin M. Murphy and Sherwin Rosen. I'd like to thank Brooks Pierce, Paul Romer, Ken Judd, Beth Ingram, Ed Prescott and Fernando Alvarez. I also thank the workshop participants of the University of Chicago, University of Pennsylvania, University of Toronto, University of Rochester, University of Washington, Penn State University, University at Buffalo, SUNY, Columbia University and University of Iowa.     

IDENTIFICATION AND ESTIMATION OF AUCTION MODEL WITH TWO‐DIMENSIONAL UNOBSERVED HETEROGENEITY*

This article investigates the empirical importance of allowing for multidimensional sources of unobserved heterogeneity in auction models with private information. It develops the estimation procedure to recover the distribution of private information in the presence of two sources of unobserved heterogeneity. It is shown that this estimation procedure identifies components of the model and produces uniformly consistent estimators of these components. The results of the estimation with highway procurement data indicate that allowing for two‐dimensional unobserved heterogeneity may significantly affect the results of estimation as well as policy‐relevant instruments derived from the estimated distributions of bidders’ costs.     

UNCERTAIN INFANT MORTALITY,LEARNING, AND LIFE‐CYCLE FERTILITY*

This article examines the links between infant mortality and fertility in an environment with unobserved heterogeneity in infant mortality risk across mothers. In such an environment, replacement behavior (i.e., the fertility response to an experienced child death) might be influenced by mothers' learning about a family‐specific component of infant mortality risk. I explicitly introduce learning by mothers in a dynamic stochastic model of life‐cycle marital fertility, and I estimate the model's structural parameters using Malaysian panel data. The framework is used to estimate replacement rates and to correct for birth selectivity in the estimation of the relationship between infant mortality risk and “health inputs.”     

A COMPUTATIONALLY PRACTICAL SIMULATION ESTIMATION ALGORITHM FOR DYNAMIC PANEL DATA MODELS WITH UNOBSERVED ENDOGENOUS STATE VARIABLES*

This article develops a simulation estimation algorithm that is particularly useful for estimating dynamic panel data models with unobserved endogenous state variables. Repeated sampling experiments on dynamic probit models with serially correlated errors indicate the estimator has good small sample properties. We apply the estimator to a model of female labor supply and show that the rarely used Polya model fits the data substantially better than the popular Markov model. The Polya model also produces far less state dependence and many fewer race effects and much stronger effects of education, young children, and husband's income on female labor supply decisions.     

An empirical note about additive outliers and nonstationarity in Latin-American inflation series

This note shows the empirical dangers of the presence of large additive outliers when testing for unit roots using standard unit root statistics. Using recent proposed procedures applied to four Latin-American inflation series, I show that the unit root hypothesis cannot be rejected.Jel classification: C2, C3, C5I want to thank Pierre Perron for useful comments on a preliminary version of this paper. Helpful comments from an anonymous referee, and Yiagadeesen Samy are appreciated. I thank the Editor Baldev Raj for useful comments about the final structure of this paper. Finally, I also thank André Lucas for helpful suggestions concerning the use of his nice computer program Robust Inference Plus Estimation (RIPE).First revision received: August 2001/Final revision received: December 2002     

Self‐employment Dynamics in Australia and the Importance of State Dependence

Using Australian labour force data, both cross‐sectional and longitudinal, this study examines the dynamics of self‐employment with a particular focus on workers transitioning between self‐employment and salaried employment, and the extent to which self‐employment is the result of workers’ observed and unobserved characteristics or is instead determined by their prior employment experience itself. Probability models of self‐employment using both pooled‐panel probit and dynamic random‐effects panel probit methods are estimated, and the results are found to be extremely sensitive to the differences in the econometric methods. Once unobserved heterogeneity and initial conditions are controlled for in the dynamic model, the importance of observed characteristics in determining self‐employment is greatly diminished. Instead, workers’ past experience in self‐employment (as opposed to salaried employment) is found to have a large favourable effect on their future self‐employment prospects. The influence of this state dependence is also considerably more important in determining self‐employment outcomes than salaried ones. Despite establishing the importance of state dependence, however, what this effect implies about why individuals choose to become self‐employed or the role that self‐employment plays in the labour market remains unresolved.     

Convenient estimators for the panel probit model: Further results

Bertschek and Lechner (1998) propose several variants of a GMM estimator based on the period specific regression functions for the panel probit model. The analysis is motivated by the complexity of maximum likelihood estimation and the possibly excessive amount of time involved in maximum simulated likelihood estimation. But, for applications of the size considered in their study, full likelihood estimation is actually straightforward, and resort to GMM estimation for convenience is unnecessary. In this note, we reconsider maximum likelihood based estimation of their panel probit model then examine some extensions which can exploit the heterogeneity contained in their panel data set. Empirical results are obtained using the data set employed in the earlier study. Helpful comments and suggestions by Irene Bertschek and Michael Lechner are gratefully acknowledged. This paper has also benefited from comments by two anonymous referees and from seminar participants at the Center for Health Economics at the University of York. Any remaining errors are the responsibility of the author.