Panel Data Models: Some Recent Developments

Arellano, Manuel; Honoré, Bo E.

doi:10.1016/s1573-4412(01)05006-1

Cited by 371 publications

(235 citation statements)

References 99 publications

Supporting

Mentioning

229

Contrasting

Unclassified

Order By: Relevance

“…Thus y i0 is not independent of x T i1 , and this dependence must be recognized when performing the integration described in the last line of equation (9). For recent discussions of this problem and further references, see Arellano and Honoré (2001) and Honoré and Tamer (2006). 16 Hsiao (1986) describes this type of approach in the context of a simpler dynamic discrete choice model.…”

Section: Estimationmentioning

confidence: 99%

Market Entry Costs, Producer Heterogeneity, and Export Dynamics

2001

View full text Add to dashboard Cite

As the exchange rate, foreign demand, production costs and export promotion policies evolve, manufacturing firms are continually faced with two issues: Whether to be an exporter, and if so, how much to export. We develop a dynamic structural model of export supply that characterizes these two decisions and estimate the model using plant-level panel data on Colombian chemical producers. The model embodies uncertainty, plant-level heterogeneity in export profits, and sunk entry costs for plants breaking into foreign markets. Our estimates, and the simulation exercises that they support, yield several implications. First, entry costs are typically large, but vary greatly across producers. Second, there is substantial cross-plant heterogeneity in gross expected export profit streams. Third, these large entry costs make expectations about future exporting conditions important for many producers, so changes in the exchange rate regime that are credible induce much more entry than those that are not. Fourth, however, most of the entry and exit takes place among marginal exporters who contribute little to aggregate export revenues. Finally, subsidies on export earnings have a much larger impact on export revenues (per dollar spent) than subsidies that reduce the entry costs faced by new exporters. (Baldwin and Krugman, 1989;Dixit, 1989;Krugman, 1989 These features of the Colombian chemical industry shape the results of our policy simulations. First, the fact that sunk entry costs are large makes expectations about future exporting conditions important for many plants. Thus we find that a moderate shift in the mean of the log exchange rate process induces sustained net entry by new exporters and rising export volumes if producers view it as a credible regime shift. On the other hand, the exact same change in the exchange rate process induces far less entry when producers retain their old beliefs about 2 Earlier studies of export market participation have focused on the null hypothesis that sunk costs don't matter, but have not been structural and thus have not quantified sunk costs (Roberts and Tybout, 1997a;Campa, 1998;Bernard and Jensen, 2001;Bernard and Wagner, 2001).3 the exchange rate process.Second, profits for the major exporters are sufficiently large to keep them in the export market under any reasonable policy scenario. So the foreign market entry and exit that takes place is concentrated among small suppliers who have a relatively minor effect on export volumes. It follows that, while expectations have a dramatic effect on the number of exporters, their effect on the volume of exports is much more modest.Third, policies that promote exports through per-unit subsidies generate far larger responses per peso spent than policies that promote exports through lump-sum transfers for new exporters. The reasons are that (1) exporters that need a subsidy to get into export markets are almost always marginal suppliers; (2) these same exporters face relatively high entry costs, and (3) large incumbent exporters, ...

show abstract

Section: Estimationmentioning

confidence: 99%

Market Entry Costs, Producer Heterogeneity, and Export Dynamics

2001

View full text Add to dashboard Cite

show abstract

“…Models and methods of controlling for unobserved heterogeneity in linear models are well established; see Chamberlain (1984) and Arellano and Honoré (2001) for references and discussion. Controlling for unobserved heterogeneity is much more difficult in nonlinear models.…”

Section: Introductionmentioning

confidence: 99%

Jackknife and analytical bias reduction for nonlinear panel models

Hahn¹,

Newey

2003

Working Paper Series

168

View full text Add to dashboard Cite

“…The literature on panel data models is large, and we do not attempt to give a complete overview here. A survey is provided by Arellano and Honoré (2001). One obvious generalization would be to allow for a non-linear relationship between y ij and x ij but to maintain additivity of the incidental function, as in…”

Section: Non-linear Specificationsmentioning

confidence: 99%

First-differencing in panel data models with incidental functions

Jochmans

2014

The Econometrics Journal

View full text Add to dashboard Cite

I discuss the fixed-effect estimation of panel data models with time-varying excess heterogeneity across cross-sectional units. These latent components are not given a parametric form. A modification to traditional first-differencing is motivated which, asymptotically, removes the permanent unobserved heterogeneity from the differenced model. Conventional estimation techniques can then be readily applied. Distribution theory for a kernel-weighted GMM estimator under large-n and fixed-T asymptotics is developed. The estimator is put to work in a series of numerical experiments to static and dynamic models.jel classification: C13, C14, C33 keywords: dynamic panel data, GMM, incidental functions, local firstdifferencing, time-varying fixed effects, nonparametric heterogeneity.

show abstract

Panel Data Models: Some Recent Developments

Cited by 371 publications

References 99 publications

Market Entry Costs, Producer Heterogeneity, and Export Dynamics

Market Entry Costs, Producer Heterogeneity, and Export Dynamics

Jackknife and analytical bias reduction for nonlinear panel models

First-differencing in panel data models with incidental functions

Contact Info

Product

Resources

About