A Generalized Non-Parametric Instrumental Variable-Control Function Approach to Estimation in Nonlinear Settings

When the endogenous variables enter non-parametrically into the regression equation standard linear instrumental variables approaches fail. Two existing solutions are the non-parametric instrumental variables (NPIVs) estimators, which are based on a set of conditional moment restrictions (CMRs), and the control function (CF) estimators, which use conditional mean independence (CMI) restrictions. Our first contribution is to show that – similar to CMI – the CMR place shape restrictions on the conditional expectation of the error given the instruments and endogenous variables that are sufficient for identification, and we call our new estimator based on these restrictions the CMR-CF estimator. Our second contribution is to develop an estimator for non-linear and non-parametric settings that can combine both CMR and CMI restrictions, which cannot be done in either the NPIV nor the non-parametric CF setting. This new “Generalized CMR-CF” uses both CMR and CMI restrictions together by allowing the conditional expectation of the structural error to depend on both instruments and control variables. When sieves are used to approximate both the structural function and the CF our estimator reduces to a series of least squares regressions. Our Monte Carlos illustrate that our new estimator performs well across several economic settings.

Entropy Balancing for Continuous Treatments

Interest in evaluating the effects of continuous treatments has been on the rise recently. To facilitate the estimation of causal effects in this setting, the present paper introduces entropy balancing for continuous treatments (EBCT) – an intuitive and user-friendly automated covariate balancing scheme – by extending the original entropy balancing methodology of Hainmueller, J. 2012. “Entropy Balancing for Causal Effects: A Multivariate Reweighting Method to Produce Balanced Samples in Observational Studies.” Political Analysis 20 (1): 25–46. In order to estimate balancing weights, the proposed approach solves a globally convex constrained optimization problem, allowing for computationally efficient software implementation. EBCT weights reliably eradicate Pearson correlations between covariates (and their transformations) and the continuous treatment variable. As uncorrelatedness may not be sufficient to guarantee consistent estimates of dose–response functions, EBCT also allows to render higher moments of the treatment variable uncorrelated with covariates to mitigate this issue. Empirical Monte-Carlo simulations suggest that treatment effect estimates using EBCT display favorable properties in terms of bias and root mean squared error, especially when balance on higher moments of the treatment variable is sought. These properties make EBCT an attractive method for the evaluation of continuous treatments. Software implementation is available for Stata and R.

Biases in Maximum Simulated Likelihood Estimation of Bivariate Models

This paper finds that the maximum simulated likelihood (MSL) estimator produces substantial biases when applied to the bivariate normal distribution. A specification of the random parameter bivariate normal model is considered, in which a direct comparison between the MSL and maximum likelihood (ML) estimators is feasible. The analysis shows that MSL produces biased results for the correlation parameter. This paper also finds that the MSL estimator is biased for the bivariate Poisson-lognormal model, developed by Munkin and Trivedi (1999. “Simulated Maximum Likelihood Estimation of Multivariate Mixed-Poisson Regression Models, with Application.” The Econometrics Journal 2: 29–48). A simulation study is conducted, which shows that MSL leads to serious inferential biases, especially large when variance parameters in the true data generating process are small. The MSL estimator produces biases in the estimated marginal effects, conditional means and probabilities of count outcomes.

Density Forecast of Financial Returns Using Decomposition and Maximum Entropy

We consider a multiplicative decomposition of the financial returns to improve the density forecasts of financial returns. The multiplicative decomposition is based on the identity that financial return is the product of its absolute value and its sign. Advantages of modeling the two components are discussed. To reduce the effect of the estimation error due to the multiplicative decomposition in estimation of the density forecast model, we impose a moment constraint that the conditional mean forecast is set to match with the sample mean. Imposing such a moment constraint operates a shrinkage and tilts the density forecast of the decomposition model to produce the improved maximum entropy density forecast. An empirical application to forecasting density of the daily stock returns demonstrates the benefits of using the decomposition and imposing the moment constraint to obtain the improved density forecast. We evaluate the density forecast by comparing the logarithmic score (LS), the quantile score (QS), and the continuous ranked probability score (CRPS). We contribute to the literature on the density forecast and the decomposition models by showing that the density forecast of the decomposition model can be improved by imposing a sensible constraint in the maximum entropy framework.

Modeling Qualitative Outcomes by Supplementing Participant Data with General Population Data: A New and More Versatile Approach

Although one often has detailed information about participants in a program, the lack of comparable information on non-participants precludes standard qualitative choice estimation. This challenge can be overcome by incorporating a supplementary sample of covariate values from the general population. This paper presents new estimators based on this sampling strategy, which perform comparably to the best existing supplementary sampling estimators. The key advantage of the new estimators is that they readily incorporate sample weights, so that they can be applied to Census surveys and other supplementary data sources that have been generated using complex sample designs. This substantially widens the range of problems that can be addressed under a supplementary sampling estimation framework. The potential for improving precision by incorporating imperfect knowledge of the population prevalence rate is also explored.

The Robustness of Conditional Logit for Binary Response Panel Data Models with Serial Correlation

We examine the conditional logit estimator for binary panel data models with unobserved heterogeneity. A key assumption used to derive the conditional logit estimator is conditional serial independence (CI), which is problematic when the underlying innovations are serially correlated. A Monte Carlo experiment suggests that the conditional logit estimator is not robust to violation of the CI assumption. We find that higher persistence and smaller time dimension both increase the magnitude of the bias in slope parameter estimates. We also compare conditional logit to unconditional logit, bias corrected unconditional logit, and pooled correlated random effects logit.

Subgraph Network Random Effects Error Components Models: Specification and Testing

This paper develops a subgraph random effects error components model for network data linear regression where the unit of observation is the node. In particular, it allows for link and triangle specific components, which serve as a basal model for modeling network effects. It then evaluates the potential effects of ignoring network effects in the estimation of the coefficients’ variance-covariance matrix. It also proposes consistent estimators of the variance components using quadratic forms and Lagrange Multiplier tests for evaluating the appropriate model of random components in networks. Monte Carlo simulations show that the tests have good performance in finite samples. It applies the proposed tests to the Call interbank market in Argentina.

Accounting for Endogeneity in Regression Models Using Copulas: A Step-by-Step Guide for Empirical Studies

We provide a detailed presentation and guide for the use of Copulas in order to account for endogeneity in linear regression models without the need for instrumental variables. We start by developing the model from first principles of likelihood inference, and then focus on the Gaussian Copula. We discuss its merits and propose diagnostics to assess its validity. We analyze in detail and provide solutions to the various issues that may arise in empirical applications for applying the method. We treat the cases of both continuous and discrete endogenous regressors. We present simulation evidence for the performance of the proposed model in finite samples, and we illustrate its application by a short empirical study. A Supplementary File contains additional simulations and another empirical illustration.

Time on the Market and Probability of Sale Using a Generalized Geometric Hazard Model

This study makes two main contributions to the applied econometrics literature. First, it shows how the all-important marginal effects for time on the market and probability of sale can be obtained from any hazard model. Second, it extends the generalization of the geometric due to Gómez-Déniz, E. 2010. “Another Generalization of the Geometric Distribution.” Test 19: 399–415 to include covariates for use in the estimation of time on the market and probability of sale regressions in real estate, thus creating an entirely new hazard model based on probability of sale rather than time on the market. For the generalized geometric we develop expressions for the marginal effects (with approximate standard errors) for both the probability of sale and time on the market. This formulation allows the impact of changes in independent variables on both the probability of sale and time on the market to be determined from a single regression model. For comparison, we also obtain these two sets of marginal effects for the popular Weibull hazard model. The geometric, generalized geometric, and Weibull hazard models, along with two sets of marginal effects for each, are estimated using data on condominium listings in a metropolitan area in Florida.