Binary Dependent Variable Models Research Articles

Separation and Rare Events

AbstractWhen separation is a problem in binary dependent variable models, many researchers use Firth's penalized maximum likelihood in order to obtain finite estimates (Firth, 1993; Zorn, 2005; Rainey, 2016). In this paper, I show that this approach can lead to inferences in the opposite direction of the separation when the number of observations are sufficiently large and both the dependent and independent variables are rare events. As large datasets with rare events are frequently used in political science, such as dyadic data measuring interstate relations, a lack of awareness of this problem may lead to inferential issues. Simulations and an empirical illustration show that the use of independent “weakly-informative” prior distributions centered at zero, for example, the Cauchy prior suggested by Gelman et al. (2008), can avoid this issue. More generally, the results caution researchers to be aware of how the choice of prior interacts with the structure of their data, when estimating models in the presence of separation.

Health status and labour market outcome: Empirical evidence from Australia

This paper uses eight waves of Australia Household, Income and Labour Dynamics data to study the issues of state dependence and the short‐run and long‐run response to health shocks on the labour market. We consider six alternative panel data binary dependent variable models with different ways of modelling labour market dynamics and individual heterogeneity. We find that the key results with regard to labour market dependence and the impacts of health shocks are sensitive to model specification and pooling of male and female samples with differences as large as sixfold. Specification analysis is conducted and favours the dynamic fixed effects logit model for separate male and female samples. Methods for evaluating dynamic response paths to a one‐time health shock for binary outcomes are also suggested and results are presented.

Health Status and Labour Market Outcome: Empirical Evidence from Australia

This paper uses eight waves of Australia Household, Income and Labour Dynamics data to study the issues of state dependence and the short‐run and long‐run response to health shocks on the labour market. We consider six alternative panel data binary dependent variable models with different ways of modelling labour market dynamics and individual heterogeneity. We find that the key results with regard to labour market dependence and the impacts of health shocks are sensitive to model specification and pooling of male and female samples with differences as large as sixfold. Specification analysis is conducted and favours the dynamic fixed effects logit model for separate male and female samples. Methods for evaluating dynamic response paths to a one‐time health shock for binary outcomes are also suggested and results are presented.

Early warning system of finance stress for India

This paper develops a composite Financial Sector Stress Index (FSSI) for the Indian financial system as a whole by combining three sub-indices for currency markets, the banking sector and the stock market, to gauge the level of financial stress in the Indian financial system. Such a continuous-valued index can be used to track the varying magnitude and dynamics of financial stress in the country over time. The FSSI provides an ordinal measure of stress in the financial system. Changes in the FSSI are useful in assessing whether financial stress is rising or falling, and in ascertaining extreme events in the financial system. The paper, then, objectively identifies extreme stress periods in the financial system based on movements of the index. We recognise that such extreme/tail events pose special econometric challenges as they are rare events but with a big impact. Hence, we use a more robust and analytically sound technique of Extreme Value Theory (EVT) to identify extreme events in the financial system. Once the extreme stress events are identified, we use a binary dependent variable model (logit model) to estimate the impact of various macroeconomic and financial variables on the probability of extreme stress in the financial system.

Analysis of factors determining the inward FDI in top seven Indian states from top seven source countries using gravity model

Using gravity model to the analysis of regional factors of FDI inflows into India, this study aims to analyse and determine the attractiveness factors of top 7 receiving Indian states from top seven investing countries' FDI for the period 2004 to 2013. Two econometrics methods were used: OLS estimation and ML-estimation of binary dependent variable models. The results suggest that gross products of host states and source countries, agglomeration effect, capital city advantages, cultural closeness and skilled labour abundance are positively related to the number of foreign firms in a particular Indian state. The distance between host states and source countries is negatively related. As for the resource abundance there is no evidence of an expected positive influence on the dependent variable. The main result of the study is that the necessary condition for FDI presence in a particular Indian states is its economic performance measured by gross regional product.

Relative Importance Analysis With Multicategory Dependent Variables:

Determining independent variable relative importance is a highly useful practice in organizational science. Whereas techniques to determine independent variable importance are available for normally distributed and binary dependent variable models, such techniques have not been extended to multicategory dependent variables (MCDVs). The current work extends previous research on binary dependent variable relative importance analysis to provide a methodology for conducting relative importance analysis on MCDV models from a dominance analysis (DA) perspective. Moreover, the current work provides a set of comprehensive data analytic examples that demonstrate how and when to use MCDV models in a DA and the advantages general DA statistics offer in interpreting MCDV model results. Moreover, the current work outlines best practices for determining independent variable relative importance for MCDVs using replicable examples on data from the publicly available General Social Survey. The present work then contributes to the literature by using in-depth data analytic examples to outline best practices in conducting relative importance analysis for MCDV models and by highlighting unique information DA results provide about MCDV models.

Industrial determinants of anti-dumping in Brazil – Protection, competition and performance: An analysis with binary dependent variable and panel data

Abstract This article investigates how sectorial characteristics determine the probability of industries receiving an anti-dumping (AD) duty, delving into the economic literature on the political economy of protection, market competition and trade reform in Brazil. By focusing on when the AD was granted, the article concentrates on the objective economic motivations for an industry to claim protection and which are instrumental to the decision of the trade defense authority. The empirical strategy involves a panel of 93 industrial sectors from 1996 to 2007 to create four groups of explanatory variables (collective action/competition, economic performance, factors of production, trade coefficients), reflecting their structural characteristics, market competition, economic performance and political mobilization. The econometric specifications apply binary dependent variable models ( logit ) with cluster robust and population-averaged error correction. The results show that structure of trade (imports volume and tariffs), economic performance (productivity and investments), and factor content (labor and natural resources) influence the probability of the AD. Political economy characteristics (captured by interactive terms) are also influential in determining the measure.

Wages and Employs for Non-Farm Agricultural Activities: One Livelihood Strategy in Nicaragua

This article analyzes the indexes for nonfarm agricultural activities, which combine agricultural activities with both employment and wage. They were made with panel data of Living Measurement Standard Survey (1993, 1998, 2001 and 2005) and they were processed with econometric model as parametric technique (Binary dependent variable model). The indexes trend was explained by a combining between nonfarm and farming agricultural activities. In summary, as soon as the economic public policy makers apply preventive measure in the labour market, while the indexes for nonfarm agricultural activities are growing up. In fact, the small farmers use the first, second and third nonfarm employ as livelihood strategy for reducing the restrictive public policy. (Unemployment).

IFRSs for financial instruments, quality of information and capital market’s volatility: an empirical assessment for Eurozone

This study examines the Eurozone stock markets’ reaction to a number of events associated with the post-implementation amendments of the IFRSs regarding financial instruments (IFRS 7, IFRS 9, IAS 32 and IAS 39). The adoption of these IFRSs is probably one of the most important recent changes in financial information environment. However, in order to contribute to an increase in financial stability, it is necessary to ensure an ex post stability of the regulatory framework. Based on this meta-argument, we have analyzed the inter-linkages between the episodes of higher market volatility and the amendments to the mentioned IFRSs, inside a binary dependent variable model. Overall, our findings suggest that the adoption of the IFRS can lead to a stabilization of European financial markets but this result is not necessary a robust one since investors prefer a higher stability on standards’ contents and application conditions.

Testing for Interaction in Binary Logit and Probit Models: Is a Product Term Essential?

Political scientists presenting binary dependent variable (BDV) models often hypothesize that variables interact to influence the probability of an event, Pr(Y). The current typical approach to testing such hypotheses is (1) estimate a logit or probit model with a product term, (2) test the hypothesis by determining whether the coefficient for this term is statistically significant, and (3) characterize the nature of any interaction detected by describing how the estimated effect of one variable on Pr(Y) varies with the value of another. This approach makes a statistically significant product term necessary to support the interaction hypothesis. We show that a statistically significant product term is neither necessary nor sufficient for variables to interact meaningfully in influencing Pr(Y). Indeed, even when a logit or probit model contains no product term, the effect of one variable on Pr(Y) may be strongly related to the value of another. We present a strategy for testing for interaction in a BDV model, including guidance on when to include a product term.

Empirical Strategies for Various Manifestations of Multilevel Data

Equivalent separate-subsample (two-step) and pooled-sample (one-step) strategies exist for any multilevel-modeling task, but their relative practicality and efficacy depend on dataset dimensions and properties and researchers' goals. Separate-subsample strategies have difficulties incorporating cross-subsample information, often crucial in time-series cross-section or panel contexts (subsamples small and/or cross-subsample information great) but less relevant in pools of independently random surveys (subsamples large; cross-sample information small). Separate-subsample estimation also complicates retrieval of macro-level-effect estimates, although they remain obtainable and may not be substantively central. Pooled-sample estimation, conversely, struggles with stochastic specifications that differ across levels (e.g., stochastic linear interactions in binary dependent-variable models). Moreover, pooled-sample estimation that models coefficient variation in a theoretically reduced manner rather than allowing each subsample coefficient vector to differ arbitrarily can suffer misspecification ills insofar as this reduced specification is lacking. Often, though, these ills are limited to inefficiencies and standard-error inaccuracies that familiar efficient (e.g., feasible generalized least squares) or consistent-standard-error estimation strategies can satisfactorily redress.

A consistent analysis of diversification decisions with non-observable firm effects

The empirical analyses of firm diversification decisions, both for new activities (new products) and markets (for example, new routes for airlines), have usually estimated a binary dependent variable model for each of the decisions the firm makes. To obtain consistent estimators, every relevant effect must be considered in the specification. As this will hardly happen, the presence of nonobserved firm effects (either because such data do not exist or because it is impossible to obtain them) must be econometrically treated, because it causes inconsistency in the estimations. In this paper we propose to use the estimators provided by the maximization of the conditional likelihood function in problems of this kind because they give consistent results even when unobserved firm effects are present. Finally, we apply this technique to an example of diversification among Spanish manufacturers. © 1997 John Wiley & Sons, Ltd.

A consistent analysis of diversification decisions with non‐observable firm effects

The empirical analyses of firm diversification decisions, both for new activities (new products) and markets (for example, new routes for airlines), have usually estimated a binary dependent variable model for each of the decisions the firm makes. To obtain consistent estimators, every relevant effect must be considered in the specification. As this will hardly happen, the presence of nonobserved firm effects (either because such data do not exist or because it is impossible to obtain them) must be econometrically treated, because it causes inconsistency in the estimations. In this paper we propose to use the estimators provided by the maximization of the conditional likelihood function in problems of this kind because they give consistent results even when unobserved firm effects are present. Finally, we apply this technique to an example of diversification among Spanish manufacturers. © 1997 John Wiley & Sons, Ltd.

On testing the logistic assumption in binary dependent variable models

This paper shows that the LM test for the validity of the logistic distribution commonly assumed in Binary Dependent Variable Models (i.e., the logit model) developed by Poirier (1980) can be obtained from a simple artificial regression. Monte Carlo simulations examine the small sample behaviour of the test statistic in comparison to the Information Matrix test for the logit model developed by Orme (1988) and Davidson and MacKinnon (1989), and two versions of the Reset test for limited dependent variable models suggested by Pagan and Vella (1989). Our results suggest that the LM test compares favourably under the null. The tests also appear to have varying power properties against different alternatives which suggests that they should all be used in investigating the validity of the logit model.

Preference Revelation for Public Goods: an Empirical Analysis

Using a micro data base of almost 1300 households, this article examines the relationship between fiscal preference (for local and provincial goods) and respondent characteristics. Since individuals were asked if expenditure on a particular functional area was sufficient, a binary dependent variable model was used to test a variety of hypotheses regarding fiscal preferences.