What Belongs Where? Variable Selection for Zero-Inflated Count Models with an Application to the Demand for Health Care

This paper develops stochastic search variable selection (SSVS) for zero-inflated count models which are commonly used in health economics. This allows for either model averaging or model selection in situations with many potential regressors. The proposed techniques are applied to a data set from Germany considering the demand for health care. A package for the free statistical software environment R is provided.

Analysing the impact of public capital stock using the NEG wage equation: a panel data approach

This paper examines the relationship between the level of public infrastructure and the level of productivity using panel data for the Spanish provinces over the period 1984-2004, a period which is particularly relevant due to the substantial changes occurring in the Spanish economy at that time. The underlying model used for the data analysis is based on the wage equation, which is one of a handful of simultaneous equations which when satisfied correspond to the short-run equilibrium of New Economic Geography theory. This is estimated using a spatial panel model with fixed time and province effects, so that unmodelled space and time constant sources of heterogeneity are eliminated. The model assumes that productivity depends on the level of educational attainment and the public capital stock endowment of each province. The results show that although changes in productivity are positively associated with changes in public investment within the same province, there is a negative relationship between productivity changes and changes in public investment in other regions.

A Panel Data Investigation of Real Exchange Rate Misalignment and Growth

The paper investigates the role of real exchange rate misalignment on long-run growth for a set of ninety countries using time series data from 1980 to 2004. We first estimate a panel data model (using fixed and random effects) for the real exchange rate, with different model specifications, in order to produce estimates of the equilibrium real exchange rate and this is then used to construct measures of real exchange rate misalignment. We also provide an alternative set of estimates of real exchange rate misalignment using panel cointegration methods. The variables used in our real exchange rate models are: real per capita GDP; net foreign assets; terms of trade and government consumption. The results for the two-step System GMM panel growth models indicate that the coefficients for real exchange rate misalignment are positive for different model specification and samples, which means that a more depreciated (appreciated) real exchange rate helps (harms) long-run growth. The estimated coefficients are higher for developing and emerging countries.

The Role of Institutions in Cross-Section Income and Panel Data Growth Models: A Deeper Investigation on the Weakness and Proliferation of Instruments

This paper investigates the role of institutions in determining per capita income levels and growth. It contributes to the empirical literature by using different variables as proxies for institutions and by developing a deeper analysis of the issues arising from the use of weak and too many instruments in per capita income and growth regressions. The cross-section estimation suggests that institutions seem to matter, regardless if they are the only explanatory variable or are combined with geographical and integration variables, although most models suffer from the issue of weak instruments. The results from the growth models provides some interesting results: there is mixed evidence on the role of institutions and such evidence is more likely to be associated with law and order and investment profile; government spending is an important policy variable; collapsing the number of instruments results in fewer significant coefficients for institutions.

International Evidence on the New Keynesian Phillips Curve Using Aggregate and Disaggregate Data

We present a unique empirical analysis of the properties of the New Keynesian Phillips Curve using an international dataset of aggregate and disaggregate sectoral in ation. Our results from panel time-series estimation clearly indicate that sectoral heterogeneity has important consequences for aggregate in ation behaviour. Heterogeneity helps to explain the overestimation of in ation persistence and underestimation of the role of marginal costs in empirical investigations of the NKPC that use aggregate data. We nd that combining disaggregate information with heterogeneous-consistent estimation techniques helps to reconcile, to a large extent, the NKPC with the data.

Regional productivity variation and the impact of public capital stock: an analysis with spatial interaction, with reference to Spain

In this paper we examine whether variations in the level of public capital across Spain‟s Provinces affected productivity levels over the period 1996-2005. The analysis is motivated by contemporary urban economics theory, involving a production function for the competitive sector of the economy („industry‟) which includes the level of composite services derived from „service‟ firms under monopolistic competition. The outcome is potentially increasing returns to scale resulting from pecuniary externalities deriving from internal increasing returns in the monopolistic competition sector. We extend the production function by also making (log) labour efficiency a function of (log) total public capital stock and (log) human capital stock, leading to a simple and empirically tractable reduced form linking productivity level to density of employment, human capital and public capital stock. The model is further extended to include technological externalities or spillovers across provinces. Using panel data methodology, we find significant elasticities for total capital stock and for human capital stock, and a significant impact for employment density. The finding that the effect of public capital is significantly different from zero, indicating that it has a direct effect even after controlling for employment density, is contrary to some of the earlier research findings which leave the question of the impact of public capital unresolved.

On decomposing the causes of changes in income-related health inequality with longitudinal data

Regression-based decomposition procedures are used to both standardise the concentration index and to determine the contribution of inequalities in the individual health determinants to the overall value of the index. The main contribution of this paper is to develop analogous procedures to decompose the income-related health mobility and health-related income mobility indices first proposed in Allanson, Gerdtham and Petrie (2010) and subsequently extended in Petrie, Allanson and Gerdtham (2010) to account for deaths. The application of the procedures is illustrated by an empirical study that uses British Household Panel Survey (BHPS) data to analyse the performance of Scotland in tackling income-related health inequalities relative to England & Wales over the five year period 1999 to 2004.

A Nonlinear Panel Unit Root Test under Cross Section Dependence

We propose a nonlinear heterogeneous panel unit root test for testing the null hypothesis of unit-roots processes against the alternative that allows a proportion of units to be generated by globally stationary ESTAR processes and a remaining non-zero proportion to be generated by unit root processes. The proposed test is simple to implement and accommodates cross sectional dependence. We show that the distribution of the test statistic is free of nuisance parameters as (N, T) −! 1. Monte Carlo simulation shows that our test holds correct size and under the hypothesis that data are generated by globally stationary ESTAR processes has a better power than the recent test proposed in Pesaran [2007]. Various applications are provided.

The Technological Specialization of Countries: An Analysis of Patent Data

New methods of analysis of patent statistics allow assessing country profiles of technological specialization for the period 1990-2006. We witness a modest decrease in levels of specialization, which we show to be negatively influenced by country size and degree of internationalization of inventive activities.

Currency Forecast Errors at Times of Low Interest Rates: Evidence from Survey Data on the Yen/Dollar Exchange Rate

Using survey expectations data and Markov-switching models, this paper evaluates the characteristics and evolution of investors' forecast errors about the yen/dollar exchange rate. Since our model is derived from the uncovered interest rate parity (UIRP) condition and our data cover a period of low interest rates, this study is also related to the forward premium puzzle and the currency carry trade strategy. We obtain the following results. First, with the same forecast horizon, exchange rate forecasts are homogeneous among different industry types, but within the same industry, exchange rate forecasts differ if the forecast time horizon is different. In particular, investors tend to undervalue the future exchange rate for long term forecast horizons; however, in the short run they tend to overvalue the future exchange rate. Second, while forecast errors are found to be partly driven by interest rate spreads, evidence against the UIRP is provided regardless of the forecasting time horizon; the forward premium puzzle becomes more significant in shorter term forecasting errors. Consistent with this finding, our coefficients on interest rate spreads provide indirect evidence of the yen carry trade over only a short term forecast horizon. Furthermore, the carry trade seems to be active when there is a clear indication that the interest rate will be low in the future.

Data-based priors for vector autoregressions with drifting coefficients

This paper proposes full-Bayes priors for time-varying parameter vector autoregressions (TVP-VARs) which are more robust and objective than existing choices proposed in the literature. We formulate the priors in a way that they allow for straightforward posterior computation, they require minimal input by the user, and they result in shrinkage posterior representations, thus, making them appropriate for models of large dimensions. A comprehensive forecasting exercise involving TVP-VARs of different dimensions establishes the usefulness of the proposed approach.

Hierarchical Shrinkage in Time-Varying Parameter Models

In this paper, we forecast EU-area inflation with many predictors using time-varying parameter models. The facts that time-varying parameter models are parameter-rich and the time span of our data is relatively short motivate a desire for shrinkage. In constant coefficient regression models, the Bayesian Lasso is gaining increasing popularity as an effective tool for achieving such shrinkage. In this paper, we develop econometric methods for using the Bayesian Lasso with time-varying parameter models. Our approach allows for the coefficient on each predictor to be: i) time varying, ii) constant over time or iii) shrunk to zero. The econometric methodology decides automatically which category each coefficient belongs in. Our empirical results indicate the benefits of such an approach.