Limited Dependent Variable Models Research Articles

SPATIAL LIMITED DEPENDENT VARIABLE MODELS: A REVIEW FOCUSED ON SPECIFICATION, ESTIMATION, AND HEALTH ECONOMICS APPLICATIONS

AbstractModeling individual choices is one of the main aim in microeconometrics. Discrete choice models have been widely used to describe economic agents' utility functions and most of them play a paramount role in applied health economics. On the other hand, spatial econometrics collects a series of econometric tools, which are particularly useful when we deal with spatially distributed data sets. Accounting for spatial dependence can avoid inconsistency problems of the commonly used statistical estimators. However, the complex structure of spatial dependence in most of the nonlinear models still precludes a large diffusion of these spatial techniques. The purpose of this paper is then twofold. The former is to review the main methodological problems and their different solutions in spatial nonlinear modeling. The latter is to review their applications to health issues, especially those appeared in the last few years, by highlighting the main reasons why spatial discrete neighboring effects should be considered and suggesting possible future lines of development in this emerging field. Particular attention has been paid to cross‐sectional spatial discrete choice modeling. However, discussions on the main methodological advancements in other spatial limited dependent variable models and spatial panel data models are also included.

Mapping the kidney disease quality of life 36-item short form survey (KDQOL-36) to the EQ-5D-3L and the EQ-5D-5L in patients undergoing dialysis

ObjectivesTo develop algorithms mapping the Kidney Disease Quality of Life 36-Item Short Form Survey (KDQOL-36) onto the 3-level EQ-5D questionnaire (EQ-5D-3L) and the 5-level EQ-5D questionnaire (EQ-5D-5L) for patients with end-stage renal disease requiring dialysis.MethodsWe used data from a cross-sectional study in Europe (France, n = 299; Germany, n = 413; Italy, n = 278; Spain, n = 225) to map onto EQ-5D-3L and data from a cross-sectional study in Singapore (n = 163) to map onto EQ-5D-5L. Direct mapping using linear regression, mixture beta regression and adjusted limited dependent variable mixture models (ALDVMMs) and response mapping using seemingly unrelated ordered probit models were performed. The KDQOL-36 subscale scores, i.e., physical component summary (PCS), mental component summary (MCS), three disease-specific subscales or their average, i.e., kidney disease component summary (KDCS), and age and sex were included as the explanatory variables. Predictive performance was assessed by mean absolute error (MAE) and root mean square error (RMSE) using 10-fold cross-validation.ResultsMixture models outperformed linear regression and response mapping. When mapping to EQ-5D-3L, the ALDVMM model was the best-performing one for France, Germany and Spain while beta regression was best for Italy. When mapping to EQ-5D-5L, the ALDVMM model also demonstrated the best predictive performance. Generally, models using KDQOL-36 subscale scores showed better fit than using the KDCS.ConclusionsThis study adds to the growing literature suggesting the better performance of the mixture models in modelling EQ-5D and produces algorithms to map the KDQOL-36 onto EQ-5D-3L (for France, Germany, Italy, and Spain) and EQ-5D-5L (for Singapore).

Mapping the Functional Independence Measure to a multi-attribute utility instrument for economic evaluations in rehabilitation: a secondary analysis of randomized controlled trial data

Purpose: To test whether the Functional Independence Measure (FIM) could be mapped to the EQ-5D-3L to give researchers a viable but “second-best” option for calculating quality-adjusted life-years (QALYs) and conducting a cost-utility analysis when only clinical outcomes have been collected.Materials and methods: Secondary analysis of repeated measures data collected during a randomized controlled trial (n = 3506 observations) at two inpatient rehabilitation centres. Participants had a mean age of 74 (SD 13) years, 63% were women and 58% were admitted with an orthopaedic diagnosis. Ordinary least-squares regression and adjusted limited dependent variable mixture models were used to estimate regression-based mappings. Performance was evaluated based on mean absolute error and the proportion of errors in excess of the minimally important difference.Results: In orthopaedic and neurological patients, high mean absolute errors (0.2 on the quality-adjusted life years scale) and a high proportion of errors (60%) in excess of the minimally important difference suggest that predicted EQ-5D-3L values provided a poor substitute for observed EQ-5D-3L values.Conclusions: Regression-based mappings from the FIM to the EQ-5D-3L are error-prone and unsuitable for calculating QALYs in rehabilitation patients. Researchers and rehabilitation professionals should therefore include a multi-attribute utility instrument such as the EQ-5D as well as the FIM to evaluate the effect of rehabilitation interventions and in rehabilitation registries. This will provide additional information on health-related quality of life and support cost-utility analyses.Implications for rehabilitationThe Functional Independence Measure (FIM) cannot be used to calculate quality-adjusted life-years (QALYs) for cost-utility analyses.Predicting QALYs from FIM data is a poor substitute for direct measurement of QALYs in orthopaedic or neurological rehabilitation populations.Multi-attribute utility instruments (MAUIs) allow direct measurement of QALYs, as well as providing a patient-reported measure of clinical quality and outcomes in rehabilitation.A MAUI should be included routinely in clinical practice by rehabilitation professionals as well as in rehabilitation trials and registries to track patient outcomes and improve clinical practice.

Development of Methods for the Mapping of Utilities Using Mixture Models: Mapping the AQLQ-S to the EQ-5D-5L and the HUI3 in Patients with Asthma

BackgroundStudies have shown that methods based on mixture models work well when mapping clinical to preference-based methods. ObjectivesTo develop these methods in different ways and to compare performance in a case study. MethodsData from 856 patients with asthma allowed mapping between the Asthma Quality of Life Questionnaire and both the five-level EuroQol five-dimensional questionnaire (EQ-5D-5L) and the health utilities index mark 3 (HUI3). Adjusted limited dependent variable mixture models and beta-based mixture models were estimated. Optional inclusion of the gap between full health and the next value as well as a mass point at the next feasible value were explored. ResultsIn all cases, model specifications formally modeling the gap between full health and the next feasible value were an improvement on those that did not. Mapping to the HUI3 required more components in the mixture models than did mapping to the EQ-5D-5L because of its uneven distribution. The optimal beta-based mixture models mapping to the HUI3 included a probability mass at the utility value adjacent to full health. This is not the case when estimating the EQ-5D-5L, because of the low proportion of observations at this point. ConclusionsBeta-based mixture models marginally outperformed adjusted limited dependent variable mixture models with the same number of components in this data set. Nevertheless, they require a larger number of parameters and longer estimation time. Both mixture model types closely fit both EQ-5D-5L and HUI data. Standard mapping approaches typically lead to biased estimates of health gain. The mixture model approaches exhibit no such bias. Both can be used with confidence in applied cost-effectiveness studies. Future mapping studies in other disease areas should consider similar methods.

Values for the ICECAP-Supportive Care Measure (ICECAP-SCM) for use in economic evaluation at end of life

End of life care may have elements of value that go beyond health. A generic measure of the benefits of end of life care could be helpful to decision makers. Such a measure, based on the capability approach, has recently been developed: the ICECAP Supportive Care Measure. This paper reports the first valuation exercise for that measure, with data from 6020 individuals collected from an on-line general population panel during June 2013. Individuals were asked to complete a stated choice experiment that combined best-worst scaling and a standard discrete choice experiment. Analysis of the best-worst data used limited dependent variable models within the random utility framework including the multinomial logit models and latent class choice model analysis. Exploratory steps were taken to determine the similarity of the best-worst and DCE data before formal testing and pooling of the two data sources. Combined data were analysed in a heteroscedastic conditional logit model adjusting for continuous scale. Two sets of tariffs were generated, one from the best-worst data capturing only main effects, and a second from the pooled data allowing for two-way interactions. Either tariff could be used in economic evaluation of interventions at the end of life, although there are advantages and disadvantages with each. This extensive valuation exercise for the ICECAP Supportive Care Measure, with a large number of members of the general public, could be complemented in the future with best-worst scaling studies amongst those experiencing the end of life.

The restricted maximum likelihood estimation of a censored regression model

It is well known in a small sample that the maximum likelihood (ML) approach for variance components in the general linear model yields estimates that are biased downward. The ML estimate of residual variance tends to be downwardly biased. The underestimation of residual variance, which has implications for the estimation of marginal effects and asymptotic standard error of estimates, seems to be more serious in some limited dependent variable models, as shown by some researchers. An alternative frequentist’s approach may be restricted or residual maximum likelihood (REML), which accounts for the loss in degrees of freedom and gives an unbiased estimate of residual variance. In this situation, the REML estimator is derived in a censored regression model. A small sample the REML is shown to provide proper inference on regression coefficients.

Determinants of households’ resistance against land acquisition for mining: Experiences at Talcher coalfields in India

In the context of people's protests against land acquisition for mining in the Indian state of Odisha, the present paper attempts to identify the underlying household level factors. The paper uses primary data collected from the Talcher coalfield region of the state and estimates limited dependent variable models. It is found that the households belonging to the scheduled tribe category or having larger per capita income or with greater access to basic amenities are more inclined to give up their land for mining. However, such inclination is lower when they have more land or better education or are concerned about deforestation and pollution. As regards the amount, the paper finds that households with higher per capita income or larger per capita asset base or greater access to basic amenities desire to give up more land. On the other hand, willingness to give up land is less either when households possess more land holding or when there is a larger male workforce in the family.

Placental thickness used as an indicator of gestational age and fetal weight in uncomplicated singleton pregnancies: A USG study in a tertiary care hospital

In the context of people's protests against land acquisition for mining in the Indian state of Odisha, the present paper attempts to identify the underlying household level factors. The paper uses primary data collected from the Talcher coalfield region of the state and estimates limited dependent variable models. It is found that the households belonging to the scheduled tribe category or having larger per capita income or with greater access to basic amenities are more inclined to give up their land for mining. However, such inclination is lower when they have more land or better education or are concerned about deforestation and pollution. As regards the amount, the paper finds that households with higher per capita income or larger per capita asset base or greater access to basic amenities desire to give up more land. On the other hand, willingness to give up land is less either when households possess more land holding or when there is a larger male workforce in the family.

Distribution analysis of train interval journey time employing the censored model with shifting character

ABSTRACTThe theoretical framework of limited dependent variable models is extended to accommodate a shifting character and thus fit the distribution of train journey time on sections of urban rail network. Data of actual train arrival and departure time at each station are used to calculate the journey time of each railway interval of multi-class trains. The log-normal distribution and normal distribution among a group of theoretical distributions are the most and second most suitable latent distributions of the train interval journey time in the censored models with shifting character. This modified distribution is described by four parameters, namely, the expectation and variance of the latent distribution and the upper and lower bound of the migration interval. The square root of the least square measurement (SRLSM) is taken as a measure, and a traversal search is adopted to determine the above four parameters according to the SRLSM. The average of the SRLSM of the theoretical train interval journey time distribution obtained by using the proposed method on all railway sections is 0.0905. The theoretical framework is the basis of storing hidden rules in data instead of past data of train travel time and optimizing the existing management of rail transit operation.

Adaptive Quadrature for Maximum Likelihood Estimation of a Class of Dynamic Latent Variable Models

Maximum likelihood estimation of models based on continuous latent variables generally requires to solve integrals that are not analytically tractable. Numerical approximations represent a possible solution to this problem. We propose to use the adaptive Gaussian---Hermite (AGH) numerical quadrature approximation for a particular class of continuous latent variable models for time-series and longitudinal data. These dynamic models are based on time-varying latent variables that follow an autoregressive process of order 1, AR(1). Two examples are the stochastic volatility models for the analysis of financial time series and the limited dependent variable models for the analysis of panel data. A comparison between the performance of AGH methods and alternative approximation methods proposed in the literature is carried out by simulation. Empirical examples are also used to illustrate the proposed approach.

Shifting the Beveridge Curve: What Affects Labor Market Matching?

This paper explores conditions and policies that could affect the matching between labor demand and supply. We identify shifts in the Beveridge curves for 12 OECD countries between 2000Q1 and 2013Q4 using three complementary methodologies and analyze the short-run determinants of these shifts by means of limited-dependent variable models. We find that labor force growth as well as employment protection legislation reduce the likelihood of an outward shift in the Beveridge curve,. Our findings also show that the matching process is more difficult the higher the share of employees with intermediate levels of education in the labor force and when long-term unemployment is more pronounced. Policies which could facilitate labor market matching include active labor market policies, such as incentives for start-up and job sharing programs. Passive labor market policies, such as unemployment benefits, as well as labor taxation render matching signficantly more difficult.

Shifting the Beveridge Curve: What Affects Labor Market Matching?

This paper explores conditions and policies that could affect the matching between labor demand and supply. We identify shifts in the Beveridge curves for 12 OECD countries between 2000Q1 and 2013Q4 using three complementary methodologies and analyze the short-run determinants of these shifts by means of limited-dependent variable models. We find that labor force growth as well as employment protection legislation reduce the likelihood of an outward shift in the Beveridge curve. Our findings also show that the matching process is more difficult the higher the share of employees with intermediate levels of education in the labor force and when long-term unemployment is more pronounced. Policies which could facilitate labor market matching include active labor market policies, such as incentives for start-up and job sharing programs. Passive labor market policies, such as unemployment benefits, as well as labor taxation render matching significantly more difficult.

Using CSR to mitigate information asymmetry in the banking sector

Abstract The paper examines the power of corporate social responsibility to reduce information asymmetry and to act as a marketing instrument in the banking sector. Trust is the most important asset of a bank. Therefore, banks are motivated to use the most effective instruments to diminish information asymmetry with their stakeholders. The fact that cash disbursements in CSR actions are not directed towards shareholders makes them more valuable signals to other stakeholders regarding the financial soundness of the bank. The empirical study conducted based on limited dependent variable models supports the effectiveness of the CSR as marketing instrument in banking. It reveals the circumstances associated to a higher probability of an active CSR policy conducted by a banking institution. The results support the hypothesis that in the banking sector CSR is perceived as an instrument which helps stakeholders reduce information asymmetry. As marketing instrument, CSR contributes to increasing the tangibility of the banking products, decreasing their perceived variability and thus making them more attractive for the clients and allowing for differentiation between competitors.

Fitting Adjusted Limited Dependent Variable Mixture Models to EQ-5D

In this article, we describe the aldvmm command for fitting adjusted limited dependent variable mixture models to either UK or U.S. tariff EQ-5D data. We present and explain the command and postestimation command through examples. The aldvmm command requires use of Stas Kolenikov's simulated annealing package (simann()), which can be easily installed by typing net install simann.pkg, from( http://web.missouri.edu/∼kolenikovs/stata ).

Environmental policies and risk finance in the green sector: Cross-country evidence

This paper provides a detailed description of venture capital investment in the green sector across 29 countries over the period 2005–2010, and identifies the role that policies might play in explaining observed cross-country differences. The analysis is based on a deal-level database of businesses seeking financing, combined with indicators of renewable policies and government R&D expenditures. The econometric analysis relates the number of deals and their volumes in a country to deployment and supply policies using count data and limited dependent variable (Tobit) models. The results suggest that both supply side policies and environmental deployment policies, designed with a long-term perspective of creating a market for environmental technologies, are associated with higher levels of venture capital relative to more short-term fiscal policies. When focusing on policies related to renewable energy generation, the results confirm the positive association of generous feed-in tariffs (FITs) with venture capital investment. However, in the solar sector excessively generous FITs tend to discourage investment, perhaps reflecting a lack of credibility over the longer term. Thus, both sets of results point to long-term policy stability, sustainability and credibility as important policy features to ensure Venture capital backing of innovative and risky ventures in a country's green sector.

Farmers' reactions to the internationalisation of cooperatives

Abstract: This study examined the extent to which farmer cooperatives enjoy social capital within their memberships as they become internationalised. The empirical basis was a survey of farmers in Finland. Findings from limited dependent variable regression models suggest that one group of farmers is loyal to the cooperative and opposes its foreign investments. A second and a third group have lower levels of social capital and switch membership status between multiple cooperatives or opt out of cooperatives. These farmers welcome international investment. This finding was more significant for individual farm types than the social capital-linked opposition to internationalisation. A general conclusion is that the internationalization of cooperatives seems affect the members’ trust in the cooperative differently. The increasing heterogeneity within the memberships may give rise to governance problems.

The Determinants of Multilateral Development Banks’ Participation in Infrastructure Projects

Multilateral development banks (MDBs) have important functions in infrastructure projects. Besides their traditional lending role, they provide additional guarantees to private investors about the creditworthiness of the project. An empirical analysis is performed for the determinants of MDBs’ participation in infrastructure projects in developing countries using data from 1990 to 2007 obtained from the World Bank’s Private Participation in Infrastructure database and through limited dependent variable models. The results suggest that the participation of MDBs is higher for less populous and poorer countries, in line with past empirical findings, and where legal and financial systems are underdeveloped. However, neither political risk nor even respect for human rights seems to play an important role in explaining multilateral participation.

A detailed decomposition for nonlinear econometric models

This paper proposes a detailed decomposition for nonlinear econometric models, covering well-known limited dependent variable models such as the logit, probit and Tobit models. The purpose of the proposed method is to relate between-group differences in an outcome variable to differences in observable characteristics. The method overcomes disadvantages of existing decompositions for nonlinear models discussed in the literature. In particular, it leads to a unique decomposition and absorbs the present nonlinearities in a natural way. In this paper, the proposed decomposition is theoretically identified and it is shown how it can be estimated and how inference can be conducted. It is then compared to existing decomposition methods. An empirical application is presented where the results from the proposed decomposition method are compared with results from existing methods. It is demonstrated that linear methods such as the Oaxaca-Blinder decomposition may not be appropriate in the presence of substantial nonlinearities, and that the decomposition approach proposed in this paper leads to a theoretically sensible decomposition of the outcome differential.

SPECIFICATION TESTING WHEN THE NULL IS NONPARAMETRIC OR SEMIPARAMETRIC

This paper discusses the problem of testing misspecifications in semiparametric regression models for a large family of econometric models under rather general conditions. We focus on two main issues that typically arise in econometrics. First, many econometric models are estimated through maximum likelihood or pseudo-ML methods like, for example, limited dependent variable or gravity models. Second, often one might not want to fully specify the null hypothesis. Instead, one would rather impose some structure like separability or monotonicity. In order to address these points we introduce an adaptive omnibus test. Special emphasis is given to practical issues like adaptive bandwidth choice, general but simple requirements on the estimates, and finite sample performance, including the resampling approximations.

Tobit or OLS? An empirical evaluation under different diary window lengths

Researchers analysing time-use data often estimate limited dependent variable models because time spent must be nonnegative and cannot be more than the total amount of time in a given observation period. While the traditional empirical technique applied to such cases is maximum likelihood estimation of a Tobit (censored regression) model, recent debate has questioned whether linear models estimated via Ordinary Least Squares (OLS) are preferable. On the one hand, Tobit models are deemed necessary to address the significant censoring (i.e. large numbers of zeroes) typically found in time-use data, in the face of which OLS estimators would be biased and inconsistent. Yet, optimization occurs over a longer period than that covered by the typical time diary (often a day), and thus some argue that reported zeroes represent a measurement problem rather than true nonparticipation in the activity, in which case OLS would be preferred. We provide direct empirical evidence on this question using the Australian Time Use Surveys, which record time-use information for two consecutive diary days, by estimating censored and linear versions of a parental child care model for both 24-hour and 48-hour windows of observation in order to determine the empirical consequences of estimation technique and diary length.