Instrumental Variable Tests Research Articles

Global Supply Chains and Cross-Border Financing

This paper provides evidence that the formation of global supply chain partnerships increases access to cross-border financing. The findings are detected in all three major financing markets - equities, syndicated loans, and public debt. Difference-in-differences and instrumental variable tests allow us to draw causal inferences from our results. We propose that our findings reflect the factor of informational visibility where less visible firms can overcome informational obstacles to cross-border financing by establishing global supply chains. Specifically, we provide evidence that firms that are small, held less by institutional investors, and followed by fewer analysts have more benefit in cross-border financing from the formation of the global supply chain. Our findings have important implications for firms attempting to integrate into the global supply- chain network. More broadly, our findings suggest that the information generated by operational activities can have important effects on subsequent financing activities.

Board diversity, firm risk, and corporate policies

We examine the effects of diversity in the board of directors on corporate policies and risk. Using a multidimensional measure, we find that greater board diversity leads to lower volatility and better performance. The lower risk levels are largely due to diverse boards adopting more persistent and less risky financial policies. However, consistent with diversity fostering more efficient (real) risk-taking, firms with greater board diversity also invest persistently more in research and development (R&D) and have more efficient innovation processes. Instrumental variable tests that exploit exogenous variation in firm access to the supply of diverse nonlocal directors indicate that these relations are causal.

Environmental governance: disentangling the relationship between economic growth and rule of law on environmental policy stringency

Economic growth and the rule of law development are two important factors that affect the stringency of environmental policy. Due to the reversible causality between the two variables, a reduced-form relationship is likely to oversimplify their effects. This paper develops a equation system to illuminate the interactive relationships between rule of law, economic development and the stringency of environmental policies and tests the robustness of the variables using sensitivity and weak instrumental variable tests. Nine determinants of rule-of-law including presence of a parliamentary system, coup d’etats, and the degree of social fractionalization, are examined. We find that economic development, the rule of law, and their interactions have direct and indirect effects on a state’s environmental policies. We further find political coups have a significant impact on adherence to the rule of law. The nonparametric simulation indicates that long-run effect of rule-of-law development on environmental protection is likely considerably stronger than its direct effect.

Board Diversity, Firm Risk, and Corporate Policies

We examine the effects of diversity in the board of directors on corporate policies and risk. Using a multi-dimensional measure, we find that greater board diversity leads to lower volatility and better performance. Lower risk levels are largely due to diverse boards adopting less risky financial policies. However, consistent with diversity fostering more efficient (real) risk-taking, firms with greater board diversity invest more in R&D and have more efficient innovation processes. Instrumental variable tests that exploit exogenous variation in firm access to the supply of diverse nonlocal directors indicate that these relations are causal.

Equity‐Based Compensation of Outside Directors and Corporate Disclosure Quality

AbstractWe examine the relationship between a firm's disclosure quality and equity‐based compensation of independent members of the board of directors. The dimensions of disclosure quality we focus on are management's earnings guidance and information flowthrough financial analysts. Using both levels and changes specifications, we find the average ratio of equity‐based pay to total pay of independent board members to be positively related to a firm's disclosure quality. Our findings are robust to the inclusion of management's equity‐based compensation, other governance measures, and financial controls, and robust to instrumental variable tests of endogeneity. Furthermore, we find directors’ equity‐based compensation to be negatively associated with the firm's cost of equity capital. Our results are consistent with equity‐based compensation providing incentives to independent directors to push for better disclosure quality.

Advertising, Attention, and Acquisition Returns

We examine the hypothesis that advertising allows a takeover target’s management to increase the firm’s profile and their own negotiating power, leading to higher subsequent takeover premiums. Our evidence from 7,095 merger bids supports this hypothesis. Moreover, we find an economically significant decrease in the acquirer’s market capitalization during the announcement period. To consider the possibility of codetermination of target advertising and takeover premiums, we employ instrumental variable tests as well as propensity matching methods and our results hold. Further, we find targets that advertise are more likely to be pursued by multiple bidders and receive revised increased bids.

Testing the Validity of the Sibling Sex Ratio Instrument

AbstractWe test the validity of the sibling sex ratio instrument suggested by Angrist and Evans using the methods proposed by Kitagawa and Huber and Mellace. The sex ratio of the first two siblings is arguably randomly assigned and influences the probability of having a third child, which makes it a candidate instrument for fertility when estimating the effect of fertility on female labor supply. However, identification hinges on the random assignment of the instrument, an instrumental exclusion restriction, and the monotonicity of fertility in the instrument. We find that the instrumental variable tests of Kitagawa and Huber and Mellace do not point to a violation of these assumptions in the Angrist and Evans data (which can, however, not be ruled out even asymptotically as the tests cannot detect all possible violations).

The Impact of Arbitrageurs on Market Liquidity

Theory predicts that arbitrage improves financial market liquidity when arbitrage opportunities arise as a result of temporary demand shocks and worsens liquidity when arbitrage opportunities arise as a result of differences in information. In this paper, I study the impact of arbitrage in Depositary Receipts (DRs) on market liquidity, using tick-by-tick data from the U.S. and five different home markets from 1996 to 2013. My analysis suggests that around 70% of the arbitrage opportunities in DRs arise as a result of demand shocks. Consistent with theory, I then show that an increase in arbitrage activity is associated with a reduction in market order imbalance and an improvement in liquidity. My results are robust to different proxies for arbitrage activity, different methodologies, and to instrumental variable tests. Overall, these findings indicate that arbitrageurs tend to trade against market order imbalance and thus enhance market integration and liquidity.

Health in the cities: When the neighborhood matters more than income

Using a rich Italian cross-sectional dataset, we estimate the effect of a neighborhood problems aggregate (including pollution, crime, and noise) on self-assessed health, presence of chronic conditions and limitations in daily activities. We address the self-selection of the residents in their neighborhoods, as well as the possible endogeneity of income with respect to health, through instrumental variable methods and several endogeneity tests. The main novelty is the sound estimation of the neighborhood effect on health using observational data, which has the advantage of providing general results that are not dependent on any experimental design. This allows us to fully compare the neighborhood effect with the traditional socioeconomic determinants of health. Our main findings are that low quality neighborhoods are strongly health damaging. This effect is comparable to the primary/upper secondary education health differential and is even higher than the impact that poor economic circumstances have on health.

Equity-Based Compensation of Outside Directors and Corporate Information Quality

We examine the relation between a firm’s disclosure quality and equity-based compensation of independent members of the board of directors. The dimensions of disclosure quality we focus on are management’s earnings guidance and information flow through financial analysts. Using both levels and changes specifications, we find the average ratio of equity-based pay to total pay of independent board members to be positively related to a firm’s disclosure quality. Our findings are robust to the inclusion of management's equity-based compensation, other governance measures, and financial controls, and robust to instrumental variable tests of endogeneity. Furthermore, we find directors' equity-based compensation to be negatively associated with the firm's cost of equity capital. Our results are consistent with equity-based compensation providing incentives to independent directors to push for better disclosure quality.

Testing for Serial Correlation by Variable Addition in Dynamic Models Estimated by Instrumental Variables

Instrumental variable tests for serial correlation can be carried out by adding lagged residuals from initial estimation to the regressors of the model under scrutiny and then checking their joint significance. It is shown that asymptotically valid tests are obtained if the lagged residuals are also added to the initial instrument set. Monte Carlo evidence suggests that useful improvements in finite sample behavior under null and alternative hypotheses can be produced when the instrument set is extended to include the relevant lagged residuals. Links with other tests are discussed and a modification allowing for conditional heteroskedasticity is described. Copyright 1994 by MIT Press.

Testing for a unit root in time series using instrumental variable estimators with pretest data based model selection

Pantula and Hall (1991) proposed instrumental variable based tests for a unit root in an ARMA( p+1, q) time series under the assumption that ( p, q) are known. In this paper we derive conditions under which the Pantula-Hall tests still converge to the Dickey-Fuller distributions when ( p, q) are chosen from the data. We propose a simple residual based statistic which can be used to estimate ( p, q) and derive conditions under which the asymptotic distribution of the ensuing unit root tests are unaffected by this method of choosing ( p, q). We also compare the local power properties of a number of unit root tests.

Some results on the finite sample significance levels of instrumental variable tests for non-nested models

Monte-Carlo estimates of the finite sample significance levels of a number of instrumental variable tests for non-nested models are provided. A simple generalization of the J-test performs better than other procedures.