Omitted Variables Bias Research Articles

The association of vertically integrated care with health care use and outcomes.

To determine whether vertically integrated hospital and skilled nursing facility (SNF) care is associated with more efficient use of postdischarge care and better outcomes. Medicare provider, beneficiary, and claims data from 2012 to 2014. We compared facility characteristics, quality of care, and health care use for hospital-based SNFs and "virtually integrated" SNFs (defined as freestanding SNFs with close referral relationships with a single hospital) relative to nonintegrated freestanding SNFs. Among patients admitted to integrated SNFs, we estimated differences in health care use and outcomes for patients originating from the parent hospital (ie, receiving vertically integrated care) versus other hospitals using linear regressions that included SNF fixed effects. We estimated bounds for our main estimates that incorporated potential omitted variables bias. We identified hospital-based SNFs based on provider data. We defined virtually integrated SNFs based on patient flows between hospitals and SNFs. We identified SNF episodes, preceding hospital stays, patient characteristics, health care use, and patient outcomes using Medicare data. Consistent with prior research, integrated SNFs performed better on quality measures and health care use relative to nonintegrated SNFs (eg, hospital-based SNFs had 11-day shorter stays compared with nonintegrated SNFs adjusting for patient characteristics, P<.001). Stroke patients admitted to hospital-based SNFs from the parent hospital had shorter preceding hospital stays (adjusted difference: -1.2days, P=.001) and shorter initial SNF stays (adjusted difference: -2.7days, P=.049); estimates were attenuated but still robust accounting for potential omitted variables bias. For stroke patients, associations between vertically integrated care and other outcomes were either statistically insignificant or not robust to accounting for potential omitted variables bias. Vertically integrated hospital and SNF care was associated with shorter hospital and SNF stays. However, there were few beneficial associations with other outcomes, suggesting limited coordination benefits from vertical integration.

Credit Quality of Trade Receivables and Its Impact on Cash Holdings and Firm Value

This paper aims to define a variable that indicates the credit quality of trade receivables and test its impact on corporate cash holdings and corporate value. Based on a sample of publicly listed firms from the United States, we provide evidence that as the credit quality of trade receivables deteriorates, firms tend to keep higher levels of cash. This finding is in line with the precautionary motive for holding cash. Additionally, when trade receivables credit quality gets worse, corporate value is reduced. In all regressions, we treat trade receivable policy variables as endogenous because of concerns regarding omitted variables bias. Moreover, we provide the first large-sample US evidence about the negative and non-linear impact of investment in trade receivables on corporate value. We utilize system GMM in all estimations. The major contribution of this study to the accounting and trade credit literature is the introduction of a variable that aims to denote the credit quality of trade receivables and the empirical evidence about the impact of deterioration in trade receivables credit quality on cash holdings and corporate value. This study also extends the literature by delivering the first large-sample evidence for the US regarding the nature of the relationship between investment in trade receivables and corporate value.

On the effect of full-fledged IT adoption on stock returns and their conditional volatility: Evidence from propensity score matching

Applied research on central banking has attempted to assess the effects of adopting full-fledged inflation targeting on several macroeconomic variables, while the impact of financial variables has been under-researched so far. This paper fills this gap by assessing and quantifying the effect of full-fledged inflation targeting on stock returns and their conditional volatility. For this purpose, it implements the Propensity Score Matching (PSM), which is immune against reverse-causality, self-selection and omitted-variable biases, on a sample of advanced countries over 1990:Q1–2015:Q1 period. The results suggest that full-fledged inflation targeting is effective in boosting stock market returns and reducing volatility.

What do tests of the relationship between employment and technical progress hide

R The debate about whether technical progress causes technological unemployment, as the Luddites argued in the early 19 th century, has recently resurfaced in the context of new technologies and automation and the so-called Fourth Industrial Revolution. We review the main issues and then consider in detail the studies of Autor and Salomons (2017, 2018). They find that after both direct and indirect effects are accounted for, technical change is, on the aggregate, employment-augmenting. They find no evidence that technical change (proxied by the growth of productivity) reduces employment growth. We demonstrate that the regressions they estimate are problematic because they approximate an accounting identity. One or two variables in the identity (output growth or both output growth and capital growth) are omitted, which implies that the coefficient of productivity growth suffers from omitted-variable bias. As the omitted variable is known, we can have a good idea of what the statistical results must be. We conclude that, unfortunately, their work does not shed light on the question they address. JEL codes : E24, O30, O47

Production Function Estimation: Biased Coefficients and Endogenous Regressors, or a Case of Collective Amnesia?

The possible endogeneity of labor and capital in production functions, and the consequent bias of the estimated elasticities, has been discussed and addressed in the literature in different ways since the 1940s. This paper revisits an argument first outlined in the 1950s, which questioned production function estimations. This argument is that output, capital, and employment are linked through a distribution accounting identity, a key point that the recent literature has overlooked. This identity can be rewritten as a form that resembles a production function (Cobb-Douglas, CES, translog). We show that this happens because the data used in empirical exercises are value (monetary) data, not physical quantities. The argument has clear predictions about the size of the factor elasticities and about what is commonly interpreted as the bias of the estimated elasticities. To test these predictions, we estimate a typical Cobb-Douglas function using five estimators and show that: (i) the identity is responsible for the fact that the elasticities must be the factor shares; (ii) the bias of the estimated elasticities (i.e., departure from the factor shares) is, in reality, caused by the omission of a term in the identity. However, unlike in the standard omitted-variable bias problem, here the omitted term is known; and (iii) the estimation method is a second-order issue. Estimation methods that theoretically deal with endogeneity, including the most recent ones, cannot solve this problem. We conclude that the use of monetary values rather than physical data poses an insoluble problem for the estimation of production functions. This is, consequently, far more serious than any supposed endogeneity problems.

Household Debt Overhang Did Hardly Cause a Larger Spending Fall During the Financial Crisis in Australia

The “debt-overhang hypothesis” – that households cut back more on their spending in a crisis when they have higher levels of outstanding mortgage debt (Dynan, 2012) – seems to be taken for granted by macroprudential authorities in several countries in their policy decisions, as well as by the international organizations that evaluate and comment on countries’ macroprudential policy. Results are presented for UK microdata that reject the debt-overhang hypothesis. The results instead support the “spending-normalization hypothesis” of Andersen, Duus, and Jensen (2016a), what can also be called the “debt-financed overspending” hypothesis – that the correlation between high pre-crisis household indebtedness and subsequent spending cuts during the crisis reflects high debt-financed spending pre-crisis and a return to normal spending during the crisis. As discussed in Svensson (2019, 2020), this is consistent with the correlation reflecting debt-financed overspending through what Muellbauer (2012) calls the “housing-collateral household demand” and Mian and Sufi (2018) the “debt-driven household demand” channel. The correlation is thus spurious and an example of omitted-variable bias. A simple model shows that consumption and debt changes are directly and strongly positively correlated, whereas consumption and debt levels are quite weakly negatively correlated. Importantly, and in contrast, examples show that there is no systematic relation between consumption cuts and levels of or changes in LTV ratios. The lack of a robust relation between consumption cuts and levels of or changes in LTV ratios implies that tests of these hypotheses should generally not be done by regressions of consumption cuts on levels of or changes in LTV ratios. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Endogeneity Bias and Other Problems in the Estimation of Production Functions

The possible endogeneity of labor and capital in production functions, and the consequent bias of the estimated elasticities, has been discussed and addressed in the literature in different ways since the 1940s. This paper revisits an argument first outlined in the 1950s, which questioned production function estimations. This argument is that output, capital and employment, are linked through a distribution accounting identity, a key point that the recent literature has overlooked. This identity can be rewritten as a form that resembles a production function (Cobb-Douglas, CES, translog). We show that this happens because the data used in empirical exercises are value (monetary) data, not physical quantities. The argument has clear predictions about the size of the factor elasticities and about what is commonly interpreted as the bias of the estimated elasticities. To test these predictions, we estimate a typical Cobb-Douglas function using five estimators and show that: (i) the identity is responsible for the fact that the elasticities must be the factor shares; (ii) the bias of the estimated elasticities (i.e., departure from the factor shares) is, in reality, caused by the omission of a term in the identity. However, unlike in the standard omitted-variable bias problem, here the omitted term is known; and (iii) the estimation method is a second-order issue. Estimation methods that theoretically deal with endogeneity, including the most recent ones, cannot solve this problem. We conclude that the use of monetary values rather than physical data poses an insoluble problem for the estimation of production functions. This is, consequently, far more serious than any supposed endogeneity problems.

What determine inflation in Pakistan: an investigation through structural equation modeling by using time series data for a period from 1975 to 2017

Interest rate is the most popular instrument of monetary policy to control inflation around the globe. It is assumed that the tight monetary policy, such as an increase in interest rate, will reduce inflation by reducing aggregate demand in the economy. However, in reality, the use of monetary tightening could be counterproductive. The monetary tightening may increase inflation dates back to 1923 when Gibson (1923) observed that correlation between interest rate and inflation is positive. There are different opinions about how the interest rate affects inflation. The most famous is called Monetary Transmission Mechanism (MTM). According to the demand side channel of MTM, increase an interest rate, therefore the aggregate demand will decrease and finally, the price level will decrease. However, the supply side channel of MTM, states that the rise of interest rate will increase the cost of production, shifting up the aggregate supply curve. This will create a rise in the general prices level. The problem in that, there are many channels of MTM which can be categorized into demand side and supply side. Many researchers take one channel and make single equation, this leads to missing variable bias. Therefore there is need to take them simultaneously which can be done by (SEM) Structural equation modeling. This research study is very significant, because this research combines both demand and supply channel simultaneously, and removes biasedness through (SEM) structural equation modeling. The Augmented Dickey-Fuller (ADF) test finds that all the variables are stationary at first difference i.e. I (1), except inflation and investment. Firstly, unrestricted structural equation model (USEM) has estimated. The mostly coefficient of USEM has found to insignificant i.e. their probability value more than 50 percent. To remove the insignificant coefficient, we are given these unnecessary coefficients the regression weight is zero for the purpose of getting significant results. Secondly, the restricted structural equation model (RSEM) is estimated. The estimated results reveal that; interest rate has a positive relation with exchange rate and industrial inputs prices, while negative relation with the price of equity and investment. Therefore, the exchange rate, price of imports, price of exports and consumption have a positive relation with inflation while investment, loan, and output have a negative relationship with inflation. Finally, the interest rate is not a significant determinant of inflation. Therefore, the exchange rate and cost channel are responsible for the transmission mechanism of interest rate to inflation in Pakistan. This study can be extended in several ways. First, this study focus on how monetary policy effect inflation, further includes in this study fiscal policy. Lastly, a regional analysis can serve as the best extension of this research. Our work was focused on Pakistan only, however a regional comparison can be appropriate.

Board gender diversity and environmental, social and governance performance of US banks: moderating role of environmental, social and corporate governance controversies

PurposeGender diversity in corporate boards is broadly studied in existing corporate governance literature. However, the role of board gender diversity on environmental, social and governance (ESG) performance of the banks is still unaccounted for. Drawing on resource dependence and legitimacy theory, this study addresses this pressing research issue. Moreover, investigation of ESG controversies as a moderator paves the existing corporate governance research to the new avenues.Design/methodology/approachData were sourced from Refinitiv database on 37 US banks from the period of 2013 to 2017. This study employs static and dynamic panel regression models that include random effects, fixed effects and dynamic generalised method of moments (GMMs) to test the hypotheses. Furthermore, system GMM is used to reduce the issue of endogeneity, measurement error, omitted variables bias and bank-specific heterogeneity.FindingsWe identify a significant positive relationship between board gender diversity and the ESG performance of US banks. However, the result propounds non-significant moderating effect of ESG controversies on the board gender diversity–ESG performance nexus.Originality/valueLiterature on board gender diversity and ESG separately and predominantly explains firm/bank's financial performance. This study is one of the pioneering attempts to explain the role of board gender diversity on ESG performance. Although incremental, however, this study also contributes to the literature on ESG in the US context.

Links between Attitudes, Mode Choice, and Travel Satisfaction: A Cross-Border Long-Commute Case Study

This paper focuses on a particular form of high mobility, namely the long journeys to work generated by cross-border job market. More precisely, it studies the impact of such behaviors on well-being by analyzing the relationships between mode choice, transport-related attitudes, socio-demographic and spatial attributes, and the level of satisfaction in the context of cross-border long commutes to Luxembourg. The statistical modelling is rooted to a conceptual framework that emphasizes the mutual dependencies between attitudes, mode choice, and satisfaction. Based on a survey among long-distance commuters (N = 3093) held in 2010 and 2011, two ordered logistic regressions, one of which including latent constructs of transport-related attitudes derived from a structural equation modelling, are developed to explain satisfaction in commuting. Main findings are: (1) Travel-related attitudes influence satisfaction with travel more than socio-demographic attributes; (2) public transport users are globally more satisfied in commuting than car drivers; (3) the socio-economic model of satisfaction is plagued by omitted variables issues; (4) the attitude model of satisfaction drops all but one socio-economic attributes (education remains) while improving adjustment (Pseudo-R-squared = 0.57 versus 0.09; BIC = 2953 versus 6059) and avoiding omitted variables bias. The effect of attitudes and other latent constructs is of paramount importance, even concealing most socio-demographic attributes to assess satisfaction. The conclusion is devoted to a discussion on the sustainability of these cross-border long commutes from the individual, social, and environmental points of view.

Health status and willingness-to-pay estimates for the benefits of improved recycling rates: evidence from Great Britain

Waste management is a challenging task around the globe. Waste disposal and recycling have important implications, not only for environmental preservation, but also for the public health, well-being, the economy and sustainable development. However, little is known about the impact of the recycling rates on public health and the willingness to pay to increase recycling. The aim of this study was to examine the relationship among household income, recycling rates and health status and to estimate the marginal willingness-to-pay (MWTP) in Great Britain. The empirical analysis relied on data from the British Household Panel Survey (BHPS) over the period 1999–2009. We estimated the impact of recycling rates and income on health status and we calculated the monetary value for a unit increase in recycling. To solve for the endogeneity issues, coming from possible reverse causality and omitted-variable bias, we implemented two instrumental variables (IV) approaches. First, we applied the Two-Stage Least Squares (2SLS) and second, we estimated a Pooled Ordered Probit model. We found that for one percent increase in recycling rates, the average MWTP was estimated between £290 and £340 per annum. Furthermore, our results show that other determinants play an additional significant role on health status, such as the employment and marital status, the age, education level and meteorological conditions. While the study provides insights about the MWTP, future studies regarding the costs of providing recycling services may offer additional useful information to help the policy makers in the decision-making process.

Firm Life Cycle and Cost of Debt

This paper examines the relation between the corporate life cycle and lending spreads. Using a sample of 20,307 firm-loan observations spanning 5,076 publicly traded U.S. firms, we find that mature firms pay lower lending spreads. This reduction is incremental to the variation explained by financial controls that may have previously been thought to sufficiently account for variation in the corporate life cycle. In our multivariate analysis, continuous measures of maturity yield point elasticities with effects as large as 1.50 bps while categorical life cycle measures indicate that firms in the introduction and decline phases pay lending spreads that are greater than firms in the mature phase (6 percent and 12 percent, respectively). We explore omitted variables bias and instrumental variable estimation in robustness testing and find these patterns persist. Our findings are consistent with theoretical predictions regarding the relationship between the corporate life cycle and various lending risks.

Maximizing donors' gifts: A comparison of actual and expected solid organ yield among VCA donors.

Vascularized Composite Allograft (VCA) transplantation provides life-changing transplants, but VCA adds complexity to the donation process and timing, possibly impeding solid organ donation. Expanding upon descriptive analyses, this study examines risk-adjusted predictions versus the observed number of organs donated by VCA donors. Our cohort included VCA donors in the United States during January 1, 2008-December 31, 2017 (n=51), using OPTN Deceased Donor Registration Form data and the Scientific Registry of Transplant Recipients (SRTR) donor yield models to calculate observed-to-expected (O:E) yield ratios. Almost all VCA donors' livers (48/51; 94.1%) and kidneys (92/102; 90.2%) were transplanted, with fewer hearts (28/51; 54.9%), lungs (46/102; 45.1%), pancreata (15/51; 29.4%), and intestines (3/51; 5.9%) transplanted. O:E ratios for overall organ yield were slightly greater than expected for VCA donors (1.10; 95% CI: 1.02-1.17). Liver (1.17: 1.08-1.27) and lung yields (1.38: 1.07-1.68) were both greater than expected, while kidney, heart, and pancreas yields were similar to expected. Across VCA types, bilateral upper limb and abdominal wall donors had better-than-expected yields while uterus, face, and unilateral upper limb donors all had similar-to-expected yields. Solid organ yield among VCA donors was as good or better than predicted, suggesting that VCA donation does not compromise recovery and transplantation of lifesaving organs.

Benefits of leisure activities for the quality of life of older South Korean adults

ABSTRACT The purpose of this study is to examine the relationship between leisure activities and an individual’s quality of life. Specifically, the study seeks to outline how engaging in leisure activities, such as those of arts and culture, as well as travelling, affects the psychological wellness of the elderly. Our study solves the omitted variables bias and time variant variables problems evident in existing research on this topic through an analytical model that combines the ‘fixed effects model’ and the ‘ordered logit model’. Thus, it offers a more accurate representation of the effects of leisure activities on the quality of life of the elderly. We find that leisure activities can improve the quality of life of the elderly and those who are restricted by financial and physical limitations. Furthermore, travelling is found to have a greater positive effect on the elderly than do arts and cultural activities. Our results suggest that these engagements must be promoted to enhance the quality of life of the elderly, which bears implications for policy makers involved in the betterment of the lives of senior citizens.

Non-Random Exposure to Exogenous Shocks: Theory and Applications

We develop new tools for estimating the causal effects of treatments or instruments that combine multiple sources of variation according to a known formula. Examples include treatments capturing spillovers in social and transportation networks, simulated instruments for policy eligibility, and shift-share instruments. We show how exogenous shocks to some, but not all, determinants of such variables can be leveraged while avoiding omitted variables bias. Our solution involves specifying counterfactual shocks that may as well have been realized and adjusting for a summary measure of non-randomness in shock exposure: the average treatment (or instrument) across such counterfactuals. We further show how to use shock counterfactuals for valid finite-sample inference, and characterize the valid instruments that are asymptotically efficient. We apply this framework to address bias when estimating employment effects of market access growth from Chinese high-speed rail construction, and to boost power when estimating coverage effects of expanded Medicaid eligibility.

The Relationship Between Subjective Well-Being and Self-Reported Health: Evidence from Ecuador

This article addresses the relationship between self-reported health and subjective well-being in two dimensions: cognitive and emotional. Using the Household Living Conditions Survey 2014, this study represents the first approach for Ecuador and Latin America to test how the two dimensions of subjective well-being explain self-reported health. The cognitive dimension is measured by a happiness question in a life-evaluative mode. Whereas the emotional dimension is proxied by an average of 16 psychosocial well-being questions that indicates how many, from the last 7 days, the person had a poor emotional state. We use descriptive statistics and a probit model with an instrumental variable approach to address the omitted variables bias and reverse causality. After controlling for socioeconomic, personal, regional, and health related variables, the results indicate that happiness or the cognitive dimension of well-being is the main predictor of self-reported health, quantitatively more important than having a recent illness (objective health measure), habits (sport) or health care (health importance). Furthermore, more days in a negative emotional state is associated with worse self-reported health.

A literature review of new methods in empirical asset pricing: omitted-variable and errors-in-variable bias

Standard procedures in empirical asset pricing suffer from various issues that are common to all regression-based methods. This work reviews recently introduced approaches that aim to mitigate problems associated with omitted factors and errors-in-variables. New methods addressing the omitted-variable bias suggest procedures for selecting appropriate control variables, aggregating the information from a large set of factors, or making existing methods robust against omitted factors. While the omitted-variable problem is present in almost all standard empirical asset pricing methods, the errors-in-variables problem is largely limited to the estimation of factor premia via two-pass regressions. New methods addressing the errors-in-variable bias implement an instrumental variable approach, suggest a generalized version of the widely used portfolio sorts procedure, or correct estimates based on an analytic expression for the bias. Ultimately, all of these new methods represent highly relevant advances for the area of empirical asset pricing, and the possibility to synthesize the most promising approaches might be worthwhile to investigate in the future.

Estimating Bargaining Power in Real Estate Pricing Models: Conceptual and Empirical Issues

The relative bargaining power of the buyer and seller is a key feature of real estate pricing models. Classic real estate studies have sought to address bargaining effects in hedonic regression models. Prior research proposes a procedure to estimate bargaining effects in hedonic regression models that depends critically on a substitution to eliminate omitted variables bias. This study shows that the proposed solution that is often cited in the real estate economics literature does not solve the omitted variables problem given that both models are merely different parameterizations of the same model, and thus produces biased estimates of bargaining power when certain property characteristics are omitted. A classic hedonic regression model of real estate prices using Corsican apartment data supports our contention, even when the assumption of bargaining power symmetry is relaxed.

Improving Estimation in Functional Linear Regression With Points of Impact: Insights Into Google AdWords

The functional linear regression model with points of impact (PoI) is a recent augmentation of the classical functional linear model with many practically important applications. In this article, however, we demonstrate that the existing data-driven procedure for estimating the parameters of this regression model can be very instable and inaccurate. The tendency to omit relevant PoI is a particularly problematic aspect resulting in omitted-variable biases. We explain the theoretical reason for this problem and propose a new sequential estimation algorithm that leads to significantly improved estimation results. Our estimation algorithm is compared with the existing estimation procedure using an in-depth simulation study. The applicability is demonstrated using data from Google AdWords, today’s most important platform for online advertisements. The R-package FunRegPoI and additional R-codes are provided in the online supplementary materials.

Gendered Well-Being: Cross-Sectional Evidence from Poor Urban Households in India

This paper is an analysis of the subjective well-being of women relative to men in low-income urban households in India. Education and employment are pathways to greater financial well-being which is presumably of great salience to poor populations. However, either or both of these may increase aspirations. The social norms that restrict the autonomy of women may mute the impact of education and employment in a way that does not happen to men. To address this hypothesis, the paper uses primary data from a survey of over 1000 respondents across slums in Delhi. The paper finds that education and employment does reduce the well-being of women relative to men. We show that these gender varying correlations are robust to omitted variables bias.