Limited Information Maximum Likelihood Research Articles

Does small firm dynamics matter for occupation diversity and job quality? Evidence from Brazil

Recent studies have explored the role of firm dynamics and the connection to job reallocation. However, limited attention has been devoted to the role of micro and small firms in this process, particularly about the quality of the jobs generated. Therefore, this article aims to measure the impact of micro and small firm dynamics on occupation diversification in Brazil, with specific emphasis on job quality. The study used data from 558 microregions within the Brazilian industrial sector between 2003 and 2015. Panel data were applied to three econometric models: Feasible Generalized Least Squares (FGLS), Driscoll–Kraay (DK) and Instrumental Variable (IV) models for robustness analysis, including Two-Stage Least Squares (2SLS), Limited Information Maximum Likelihood (LIML), and the Generalized Method of Moments with Continuously Updating Estimators (GMM-CUE). Our main finding reveals that the dynamics of micro and small firms positively impact occupation diversity in Brazil, leading to the creation of a wider range of job types. Furthermore, the frequency of change of firms from microenterprises to small businesses increases the occupation diversity in the Brazilian industrial sector. Our findings are significant in providing policy recommendations for developing countries to achieve a more diverse labor market.

Comparison of instrumental variable methods with continuous exposure and binary outcome: A simulation study.

Instrumental variable (IV) methods are widely employed to estimate causal effects when concerns regarding unmeasured confounders. Although comparisons among several IV methods for binary outcomes exist, comprehensive evaluations are insufficient. Therefore, in this study, we aimed to conduct a simulation with some settings for a detailed comparison of these methods, focusing on scenarios where IVs are valid and under effect homogeneity with different instrument strengths. We compared six IV methods under 32 simulation scenarios: two-stage least squares (2SLS), two-stage predictor substitutions (2SPS), two-stage residual inclusions (2SRI), limited information maximum likelihood (LIML), inverse-variance weighted methods with a linear outcome model (IVWLI), and inverse-variance weighted methods with a non-linear model (IVWLL). By comparing these methods, we examined three key estimates: the parameter estimates of the exposure variable, the causal risk ratio, and the causal risk differences. Based on the results, six IV methods could be classified into three groups: 2SLS and IVWLI, 2SRI and 2SPS, and LIML and IVWLL. The first pair showed a clear bias owing to outcome model misspecification. The second pair showed a relatively good performance when strong IVs are available; however, the estimates suffered from a significant bias when only weak IVs are used. The third pair produced relatively conservative results, although they were less affected by weak IV issues. The findings indicate that no panacea is available for the bias associated with IV methods. We suggest using multiple IV methods: one for primary analysis and another for sensitivity analysis.

Approaches to estimate bidirectional causal effects using Mendelian randomization with application to body mass index and fasting glucose.

Mendelian randomization (MR) is an epidemiological framework using genetic variants as instrumental variables (IVs) to examine the causal effect of exposures on outcomes. Statistical methods based on unidirectional MR (UMR) are widely used to estimate the causal effects of exposures on outcomes in observational studies. To estimate the bidirectional causal effects between two phenotypes, investigators have naively applied UMR methods separately in each direction. However, bidirectional causal effects between two phenotypes create a feedback loop that biases the estimation when UMR methods are naively applied. To overcome this limitation, we proposed two novel approaches to estimate bidirectional causal effects using MR: BiRatio and BiLIML, which are extensions of the standard ratio, and limited information maximum likelihood (LIML) methods, respectively. We compared the performance of the two proposed methods with the naive application of UMR methods through extensive simulations of several scenarios involving varying numbers of strong and weak IVs. Our simulation results showed that when multiple strong IVs are used, the proposed methods provided accurate bidirectional causal effect estimation in terms of median absolute bias and relative median absolute bias. Furthermore, compared to the BiRatio method, the BiLIML method provided a more accurate estimation of causal effects when weak IVs were used. Therefore, based on our simulations, we concluded that the BiLIML should be used for bidirectional causal effect estimation. We applied the proposed methods to investigate the potential bidirectional relationship between obesity and diabetes using the data from the Multi-Ethnic Study of Atherosclerosis cohort. We used body mass index (BMI) and fasting glucose (FG) as measures of obesity and type 2 diabetes, respectively. Our results from the BiLIML method revealed the bidirectional causal relationship between BMI and FG in across all racial populations. Specifically, in the White/Caucasian population, a 1 kg/m2 increase in BMI increased FG by 0.70 mg/dL (95% confidence interval [CI]: 0.3517-1.0489; p = 8.43×10-5), and 1 mg/dL increase in FG increased BMI by 0.10 kg/m2 (95% CI: 0.0441-0.1640; p = 6.79×10-4). Our study provides novel findings and quantifies the effect sizes of the bidirectional causal relationship between BMI and FG. However, further studies are needed to understand the biological and functional mechanisms underlying the bidirectional pathway.

Bayesian Estimation of Simultaneous Equation Models with Outliers and Multicollinearity Problem

Outliers and multicollinearity are problems in the analysis of Simultaneous Equation Model (SEM) especially in applied research. They can lead to bias or inefficiency of estimators. This study employed a Bayesian technique for estimation of SEM that is characterized with both multicollinearity and outliers. Monte Carlo experiment was applied while the data sets with specified outliers and multicollinearity were simulated for the SEM. The estimates of Bayesian and classical methods namely; two stage Least Squares (2SLS), Three Stage Least Squares (3SLS), Limited Information Maximum Likelihood (LIML) and Ordinary Least Squares (OLS) in simultaneous equation model were then compared. The criteria used for comparison were the Mean Square Error (MSE) and Absolute Bias (ABIAS). The Bayesian method of estimation outperformed classical methods in terms of MSE and ABIAS. However, the classical method has the same performance with Bayesian method when there are no outliers and multicollinearity in the simultaneous equation model. Hence, Bayesian method of estimation is preferred than classical method when there is problem of outlier and multicollinearity in a just identified simultaneous equation model.

ICT, Financial Inclusion and the Informal Economy: Evidence from Selected Latin American and Caribbean Countries

ABSTRACT Exploring effective tools to control the rampant expansion of the informal economy is critical to achieving the Sustainable Development Goals (SDGs). Based on balanced panel data for 19 selected Latin American and Caribbean countries for the period 2008–2017, this study employs fixed-effects estimation, limited information maximum likelihood estimation, and system generalized method of moments estimation to investigate the direct and indirect effects of information and communication technologies (ICT) diffusion on the size of the informal economy. The findings suggest that ICT diffusion can significantly reduce the size of the informal economy in LAC. Moreover, the dampening effect of ICT on the informal economy is more significant in countries with lower levels of financial inclusion. Accordingly, we recommend strengthening ICT sector development to enhance financial deepening and government efficiency and thus reduce informality in LAC.

A weighted average limited information maximum likelihood estimator

In this article, a Stein-type weighted limited information maximum likelihood (LIML) estimator is proposed. It is based on a weighted average of the ordinary least squares (OLS) and LIML estimators, with weights inversely proportional to the Hausman test statistic. The asymptotic distribution of the proposed estimator is derived by means of local-to-exogenous asymptotic theory. In addition, the asymptotic risk of the Stein-type LIML estimator is calculated, and it is shown that the risk is strictly smaller than the risk of the LIML under certain conditions. A Monte Carlo simulation and an empirical application of a green patent dataset from Nordic countries are used to demonstrate the superiority of the Stein-type LIML estimator to the OLS, two-stage least squares, LIML and combined estimators when the number of instruments is large.

Digital finance and performance in innovation: Evidence from universities.

This study evaluated 72 universities' performance innovation during 2011 to 2019 of panel data, using the data envelopment analysis-Malmquist method. The study used benchmark regression to analyse the relationship between digital finance and the universities' innovation performance. The aim was to improve innovation performance and promote national innovation across countries. According to the results of the empirical analysis, digital finance positively affects innovation performance. That finding was confirmed through advanced robustness test evaluation, such as limited information maximum likelihood, two-stage least squares, and interactive fixed effects. Moreover, based on information theory, the digital finance influence mechanism improves credit demand and financial efficiency. Additionally, innovation performance survived spatial overflow effects. Lastly, the paper concludes with some implications for improving digital financial coverage and constructing innovation networks among universities.

Impact of institutional quality on ecological footprint: New insights from G20 countries

The complex relationship between institutional quality and ecological footprint (ECF) has been verified in prior literature; however, the subject is still inadequately explored and often disregarded to highlight areas where specific policy tensions exist. Therefore, this study aims to delve into the significance of institutional quality on ECF through a limited information maximum likelihood (LIML) panel analysis in the G20 countries from 2000 to 2022. For a precise evaluation, the study innovatively constructs a composite institutional quality index (IQI) and three indices under the accountability, transparency, and participation dimensions. Additionally, the study develops an economic uncertainty predictor to capture the effects of external economic shocks on the subject. The results obtained from the LIML estimation demonstrate that IQI is substantive in reducing ECF in the recipient panel. Further, the findings highlight that all three dimensions of IQI are significant in abating ECF, while the transparency index yields a higher influence on abating ECF. It is also observed that IQI plays an effective role in modulating the relationships between ECF, financial development, human development, growth, and energy consumption, while it is found to be insignificant in reducing the negative effects of globalization on ECF. Finally, the results indicate that IQI is highly influential in eliminating the adverse impact of external shocks on ECF. The results are robust and have specific policy implications.

The impact of fiscal decentralization on economic growth: A comparative analysis of selected African and OECD countries

Fiscal decentralization, widely believed to be one essential tool to boost economic growth, is typically presented as something that works for developed countries. However, this needs to be investigated, which is what this paper aims to do. This paper investigates the relationship between fiscal decentralization and economic growth, specifically focusing on its applicability in both developed and developing countries. Using cross-sectional data from 23 African and 23 OECD countries, we employ two-stage estimation methods, including two-stage least squares (2SLS), Generalized Method of Moments (GMM), and Limited Information Maximum Likelihood (LIML), to address concerns related to endogeneity. The inclusion of instruments representing country size, ethnoreligious diversity, and administrative structure enhances the analysis. The dependent variable is based on average values for the five-year period between 2015 and 2019, while the explanatory variables are derived from data in 2017, representing the midpoint of the 2013–2021 period. The results highlight the effectiveness of the 2SLS method in estimating the relationship between fiscal decentralization, control variables, instrument variables, and per capita GDP. Empirical findings indicate significant positive impacts of both expenditure decentralization and revenue decentralization on per capita GDP, holding true for both developed and developing countries, with a slightly stronger effect observed in the latter. These results underscore the potential benefits of fiscal decentralization across diverse economies and offer valuable insights for policymakers.The study uses cross-sectional data from 23 African and 23 OECD countries and applies two-stage estimation methods to effectively address the endogeneity issue in the modeling. A variety of such methods in two-stage least squares (2SLS), Generalized Method of Moments (GMM), and Limited Information Maximum Likelihood (LIML) are applied and compared against the ordinary least squares (OLS) method. Variables that represent the size, the ethnoreligious diversity and the type of administrative structure of the countries are employed as instruments. The 2SLS method proved to excellently estimate the relationship between the fiscal decentralization terms as well as other control and instrument variables and the per capita GDP. The findings from the empirical analysis showed that both expenditure decentralization and revenue decentralization have significant positive impacts on per capita GDP of a country. This relationship holds not only for developed countries but also for developing ones, even to a slightly better degree in the latter.

Agricultural machinery service adoption and farmland transfer-in decision: evidence from rural China

With the deepening of China’s agricultural labor division, the rapid development of agricultural machinery service has had a profound impact on farmland scale management. Based on the new classical economic and transaction cost theories, this paper investigates the impact mechanism of agricultural machinery service on farmers’ farmland transfer-in decision from the dual perspectives of transaction cost and benefit of labor division, and uses the national rural survey data on the Chinese Family Database to conduct empirical research by using the ordinary least square (OLS) model and limited information maximum likelihood (LIML) method. The study finds that, first, agricultural machinery service expands farmers’ farmland transfer-in decision by encouraging them to obtain the economic efficiency of labor division. Not only agricultural machinery service can alleviate the constraint of farmers’ farmland transfer in labor and capital but also reduce the service transaction cost and increase farmers’ share of the division of labor economy, which further stimulates farmers to expand the scale of management. Second, with the deepening of labor division of agricultural machinery services, the higher transaction efficiency of agricultural machinery service promotes farmers’ farmland transfer-in participation. As long as the marginal net benefit obtained is greater than zero, farmers have incentives to expand the scale of farmland. Therefore, in order to promote China’s farmland scale management, the professional, whole-process, and large-scale development of agricultural machinery services should be further promoted, and the service capacity and fields should be expanded so that outsourcing services can be better provided to agricultural production demanders. At the same time, the technical standards, supervision system, and service platform of agricultural machinery should be established and improved so as to enhance the bargaining position of farmers in service outsourcing transactions, thus promoting the development of agricultural division of labor to a higher level and realizing the mutual promotion and mutual benefit of agricultural machinery service subjects and agricultural land scale management subjects.

R&D contribution in TFP growth of Greek industry: A limited information likelihood approach

ABSTRACTWe provide a novel panel model to decompose total factor productivity (TFP) growth in the Greek industry at the firm level while we tackle the contribution of R&D. We, therefore, opt for parametric methodology that provides statistical inference and would validate the results. Our modeling departs from prior strong assumptions such as error terms across firms being independent. In fact, we provide a novel limited information maximum likelihood (LIML) estimation method that adequately deals with the issue of endogeneity and model misspecification. We demonstrate that our model detects variability in terms of TFP growth components across industries and firms. Our results show that R&D would enhance TFP of Greek firms, albeit the crisis has had a detrimental impact. Financial ratios such as liquidity and solvency ratios also affect TFP as we demonstrate that both would enhance TFP. The solvency ratio is important as it provides an estimate of whether the firm can cope with debt. We also note variability across small versus medium and large firms and report that small firms are more productive and spend more of their revenues on R&D. In terms of policy, our evidence warrants higher R&D spending to enhance TFP growth, though R&D funding is a concern.

COVID-19 Policies, Pandemic Disruptions, and Changes in Child Mental Health and Sleep in the United States

The adverse effects of COVID-19 containment policies disrupting child mental health and sleep have been debated. However, few current estimates correct biases of these potential effects. To determine whether financial and school disruptions related to COVID-19 containment policies and unemployment rates were separately associated with perceived stress, sadness, positive affect, COVID-19-related worry, and sleep. This cohort study was based on the Adolescent Brain Cognitive Development Study COVID-19 Rapid Response Release and used data collected 5 times between May and December 2020. Indexes of state-level COVID-19 policies (restrictive, supportive) and county-level unemployment rates were used to plausibly address confounding biases through 2-stage limited information maximum likelihood instrumental variables analyses. Data from 6030 US children aged 10 to 13 years were included. Data analysis was conducted from May 2021 to January 2023. Policy-induced financial disruptions (lost wages or work due to COVID-19 economic impact); policy-induced school disruptions (switches to online or partial in-person schooling). Perceived stress scale, National Institutes of Health (NIH)-Toolbox sadness, NIH-Toolbox positive affect, COVID-19-related worry, and sleep (latency, inertia, duration). In this study, 6030 children were included in the mental health sample (weighted median [IQR] age, 13 [12-13] years; 2947 [48.9%] females, 273 [4.5%] Asian children, 461 [7.6%] Black children, 1167 [19.4%] Hispanic children, 3783 [62.7%] White children, 347 [5.7%] children of other or multiracial ethnicity). After imputing missing data, experiencing financial disruption was associated with a 205.2% [95% CI, 52.9%-509.0%] increase in stress, a 112.1% [95% CI, 22.2%-268.1%] increase in sadness, 32.9% [95% CI, 3.5%-53.4%] decrease in positive affect, and a 73.9 [95% CI, 13.2-134.7] percentage-point increase in moderate-to-extreme COVID-19-related worry. There was no association between school disruption and mental health. Neither school disruption nor financial disruption were associated with sleep. To our knowledge, this study presents the first bias-corrected estimates linking COVID-19 policy-related financial disruptions with child mental health outcomes. School disruptions did not affect indices of children's mental health. These findings suggest public policy should consider the economic impact on families due to pandemic containment measures, in part to protect child mental health until vaccines and antiviral drugs become available.

HIGHER-ORDER APPROXIMATION OF IV ESTIMATORS WITH INVALID INSTRUMENTS

This paper analyzes the higher-order approximation of instrumental variable (IV) estimators in a linear homoskedastic IV regression model when a large set of instruments with potential invalidity is present. We establish theoretical results on the higher-order mean-squared error (MSE) approximation of the two-stage least-squares (2SLS), the limited information maximum likelihood (LIML), the Fuller (FULL), the bias-adjusted 2SLS, and jackknife version of the LIML and FULL estimators by allowing for local violations of the instrument exogeneity conditions. Based on the approximation to the higher-order MSE, we consider the instrument selection criteria that can be used to choose among the set of available instruments. We demonstrate the asymptotic optimality of the instrument selection procedure proposed by Donald and Newey (2001, Econometrica 69, 1161–1191) in the presence of locally (faster than $N^{-1/2}$ ) invalid instruments in the sense that the dominant term in the MSE with the chosen instrument is asymptotically equivalent to the infeasible optimum. Furthermore, we propose instrument selection procedures to choose instruments among the sets of conservative (known) valid instruments and potentially locally ( $N^{-1/2}$ ) invalid instruments based on the higher-order MSE of the IV estimators by considering the bias-variance trade-off.

Human capital and the middle-income trap revisited

ABSTRACT Middle-income trap refers to the economic growth strategies that transition low-income countries into middle-income ones but fail to transition the middle-income countries into high-income countries. We observe the existence of a middle-income trap for upper-middle- and lower-middle-income countries. We examine the reasons for the middle-income trap using the Bayesian model averaging (BMA) and generalized method of moments (GMM). We also explore the transformation of middle-income economies into high-income economies using logistic, probit and Limited Information Maximum Likelihood (LIML) regression analyses. Random forest analysis is also used to check the robustness of the findings. BMA analysis shows that education plays an enabling role in high-income countries in determining economic growth, whereas the full potential of education is not fully utilized in middle-income countries. GMM estimations show that the education coefficient is positive and significant for high-income and middle-income countries. This implies that education plays a decisive positive role in achieving economic growth and gives a path to escape from the middle-income trap. However, the education coefficient for middle-income countries is approximately half that of high-income countries. Therefore, the findings of this study call for additional investment and focused strategies relating to human capital endowments.

Bank Capital, Risk and Profitability: A Comparative Study of Europe and Mena Countries

The purpose of this paper is to investigate the role of capital regulations in relation to bank risk and profitability across two different regions i.e. Europe and the Middle East and North Africa (MENA). The sample for this study includes annual data of 502 banks from European and MENA banking sectors for the period 2010-2019. The study employs OLS regression to ascertain whether Basel and non-Basel based capital ratios affect bank risk and profitability across different economic regions. The study also used Generalized Method of Moments and Limited Information Maximum Likelihood to remove the issue of possible endogeneity problems. The results show bank capital ratios comprises of risk-weighted assets are more effective in minimizing credit default risk and increasing profitability of banks in the European and MENA regions. Moreover, findings imply that complying with Basel capital guidelines do not penalize bank activities in one region compared to another. Overall results imply that bank regulators in both regions should make policies in compliance with Basel (III) capital guidelines to achieve financial stability.

Simultaneous Maximum Likelihood Estimation for Piecewise Linear Instrumental Variable Models.

Analysis of instrumental variables is an effective approach to dealing with endogenous variables and unmeasured confounding issue in causal inference. We propose using the piecewise linear model to fit the relationship between the continuous instrumental variable and the continuous explanatory variable, as well as the relationship between the continuous explanatory variable and the outcome variable, which generalizes the traditional linear instrumental variable models. The two-stage least square and limited information maximum likelihood methods are used for the simultaneous estimation of the regression coefficients and the threshold parameters. Furthermore, we study the limiting distribution of the estimators in the correctly specified and misspecified models and provide a robust estimation of the variance-covariance matrix. We illustrate the finite sample properties of the estimation in terms of the Monte Carlo biases, standard errors, and coverage probabilities via the simulated data. Our proposed model is applied to an education-salary data, which investigates the causal effect of children’s years of schooling on estimated hourly wage with father’s years of schooling as the instrumental variable.

Environmental sustainability risk, institutional effectiveness and urbanization

This study examines the effect of urbanization, institutional quality and other variables on environmental sustainability risk among economies in Sub-Saharan Africa (SSA). Environmental sustainability risk is examined from three main perspectives – emission of gases inimical to the environment (CO2 emissions), exploitation of land and water resources (ecological footprint) and exploitation of resources devoid of the requisite savings and education expenditure (genuine sustainability risk). Empirical analyses were performed using the Limited Information Maximum Likelihood (LIML) Instrumental Variable estimation technique. Estimated results suggest that urbanization, population growth and increase commodity prices on the global market (for key export commodities) worsen CO2 emissions among economies in the sub-region. The results further suggest that urbanization, population growth, governance and institutional structures increase resource exploitation. Additional results suggest that institutional quality lessens adverse effect urbanization has on environmental sustainability risk (CO2 emissions) and genuine sustainability risk; but exacerbates the impact urbanization has on ecological footprint. Given these findings, this study recommends conscious efforts at improving governance and institutional structures among governments in the sub-region to lessen adverse effect of urbanization and population growth on environmental sustainability risk.

The Effect of Health on the Elderly's Labor Supply in Rural China: Simultaneous Equation Models With Binary, Ordered, and Censored Variables.

In this study, we examined the effect of health on the elderly's labor supply in rural China based on the data of the Chinese Health and Nutrition Survey (CHNS) from 1997 to 2006. We used simultaneous equations to address the endogeneity problem of health and estimate the models with censored data of labor supply by the full information maximum likelihood estimation. We found that the failing health does not significantly decrease the elderly's labor supply in rural areas when using both the subjective (self-reported health status) and objective (hypertension diagnosed or not) health indicators. Our finding indicates the phenomenon of “ceaseless toil” for the elderly in rural China, i.e., the elderly almost work their whole life even if they are not physically capable. The results remain robust when using a two-stage limited information maximum likelihood estimation.

Financial development and institutional quality among emerging economies

PurposeThis study examines how institutional quality influences variability in financial development among economies in Sub-Saharan Africa (SSA).Design/methodology/approachEmpirical estimations verifying various relationships are performed using the limited information maximum likelihood (LIML) estimation technique.FindingsThe results suggest that institutional quality enhances the pace of financial development among economies in the sub-region all things being equal. In a further micro-level analysis where components of institutional quality index are examined separately, the study’s results suggest that effective governance, regulatory quality, rule of law and accountability tend to have a significant positive impact on financial sector development.Research limitations/implicationsFindings of the study suggest that policies geared towards improving governance and regulatory institutions can augment development of the financial sector among economies in SSA; governments and policymakers are therefore encouraged to resource noted institutions to play effective roles for the development of the financial sector.Originality/valueCompared to related studies, this study reorients existing paradigm, which emphasizes the role of governance and institutional variables in the economic growth discourse. The authors’ empirical inquiry rather focuses on how governance and institutional structures influence regional financial development dynamics. Specifically, this study differs from most macro-level studies found in literature because it examines the impact of hitherto unexamined governance and institutional variables on financial development among economies in SSA.

Machine Labor

The utility of machine learning (ML) for regression-based causal inference is illustrated by using lasso to select control variables for estimates of college characteristics’ wage effects. Post-double-selection lasso offers a path to data-driven sensitivity analysis. ML also seems useful for an instrumental variables (IV) first stage, since two-stage least squares (2SLS) bias reflects overfitting. While ML-based instrument selection can improve on 2SLS, split-sample IV and limited information maximum likelihood do better. Finally, we use ML to choose IV controls. Here, ML creates artificial exclusion restrictions, generating spurious findings. On balance, ML seems ill-suited to IV applications in labor economics.