An unsettled issue in the literature relating to the relative wage effect of unions is the appropriate treatment of union status in a wage determination model. In the context of a three-equation model determining union membership and union- and nonunion-sector wage rates, this paper presents an instrumental variables (IV) procedure for estimating the parameters of the wage equations and a test of the exogeneity of union status using the Hausman specification test. An advantage of our IV procedure in comparison to the widely used inverse Mill's ratio procedure is that our procedure is a distribution-free estimator, whereas the inverse Mill's ratio estimator hinges in the assumption that the error term of the choice equation is normally distributed. Using data for a sample of middle-aged white workers, we estimate the parameters of the union and nonunion wage equations with both procedures. On the key question of the endogeneity of union status, the Hausman test decisively rejects the null hypothesis of exogeneity. The inverse Mill's ratio procedure, in contrast, provides coefficient estimates on the selectivity terms that fail to indicate evidence of sample selectivity in either sector.

We construct novel measures of financial constraints using textual analysis of firms’ annual reports and investigate their impact on stock returns. Our three measures capture access to equity markets, debt markets, and external financial markets in general. In all cases, constrained firms earn higher returns, which move together and cannot be explained by the Fama and French (2015) factor model. A trading strategy based on financial constraints is most profitable for large, liquid stocks. Our results are strongest when we consider debt constraints. A portfolio based on this measure earns an annualized risk-adjusted excess return of 6.5%. Received April 4, 2016; editorial decision December 17, 2017 by Editor Andrew Karolyi.

The expectations theory of the term structure is well known to give wrong signals as to the future course of long-term interest rates. One explanation involves rational time-varying term premia. However, the “anomaly” may also be due to inflation forecast errors. We study survey forecasts of inflation. It seems that the respondents' forecasts are insufficiently adaptive. Interest rates reflect expectations similar to the inflation forecasts. As a result, past survey forecast errors reliably predict premia on U.S. Government Bonds.

We propose a theory of corporate liquidity management in which signaling through illiquidity is cheaper than signaling through “skin in the game.” This causes ex post liquidation of worthy projects even when there are enough aggregate resources available for their continuation. We examine the policy remedies for this distortion and consider their consequences for the supply of liquidity, interest rate policy, and subsidies.

Weak bank supervision could give banks the ability to shift risk from themselves to supervisors. We use cross-border bank mergers as a natural experiment to test changes in risk and the impact of supervision. We examine cross-border bank mergers and find that the supervisory structures of the partners’ countries influence changes in post-merger total risk. An acquirer from a country with strong supervision lowers total risk after a cross-border merger. However, total risk increases when the target bank is located in a country with relatively strong supervision. This result is consistent with strong host regulators limiting the risky activities of their local banks. Foreign-owned competitors could then engage in the risky projects, especially if the foreign banks’ supervisors are not strong. An acquirer entering a country with strong supervision appears to shift risk back to its home country. The results suggest that bank supervisors can reduce total banking risk in their countries by being strong.

The audit fee research literature argues that auditors' costs of developing brand name reputations, including top‐tier designation and recognition for industry specialization, are compensated through audit fee premiums. Audited firms reduce agency costs by engaging high‐quality auditors who monitor the levels and reporting of discretionary expenditures and accruals. In this study we examine whether specialist auditor choice is associated with a particular discretionary expenditure ‐ research and development (R&D). For a large sample of U.S. companies from a range of industries, we find strong evidence that R&D intensity is positively associated with firms' choices of auditors who specialize in auditing R&D contracts. Additionally, we find that R&D intensive firms tend to appoint top‐tier auditors. We use simultaneous equations to control for interrelationships between dependent variables in addition to single‐equation ordinary least squares (OLS) and logistic regression models. Our results are particularly strong in tests using samples of small firms whose auditor choice is not constrained by the need to appoint a top‐tier auditor to ensure the auditor's financial independence from the client.

This paper examines how analysts combine earnings and dividend information when they predict future earnings. Because both earnings and dividends are noisy indications of future earnings, we posit that analysts use the two corroboratively, to confirm the information reflected in each, and that analysts will substitute away from earnings when it is noisy and toward dividends. Using regressions of analysts' earnings forecast revisions on unexpected earnings, unexpected dividends, and five variables that reflect whether the signs of unexpected earnings and dividends confirm or contradict each other, we find evidence of both corroboration and substitution. Analysts' earnings forecast revisions are significantly related to the five corroborative variables, and this relation has statistically significant explanatory power beyond that in the magnitudes of unexpected earnings and unexpected dividends. Consistent with expectations, we find that the evidence of corroboration varies across the noisiness of earnings information; there is more evidence of corroboration when earnings are more variable. We also find evidence consistent with analysts substituting away from earnings, toward dividend information for firms with noisy earnings information (high variance). Overall, the results imply that analysts use earnings and dividend information interdependently, with some interdependency determined by the noisiness of earnings announcements. Résumé. Les auteurs examinent comment les analystes combinent l'information relative aux bénéfices et aux dividendes pour prévoir les bénéfices futurs. Les bénéfices et les dividendes étant tous deux des indicateurs imparfaits des bénéfices futurs, les auteurs posent l'hypothèse que les analystes utilisent les deux, à titre corroboratif, pour confirmer l'information que livre chacun de ces indicateurs et qu'ils préféreront les dividendes aux bénéfices, si ces derniers se révèlent un indicateur imparfait. En procédant à la régression des révisions des prévisions de bénéfices des analystes sur les bénéfices imprévus, sur les dividendes imprévus et sur cinq variables indiquant si les pronostics de bénéfices et de dividendes imprévus se confirment ou s'infirment les uns les autres, les auteurs enregistrent des données qui vont à la fois dans le sens de la corroboration et de la substitution. Les révisions des prévisions de bénéfices des analystes présentent une relation significative avec les cinq variables de corroboration, relation qui affiche un pouvoir d'explication statistiquement significatif, au‐delà de celui de l'ampleur des bénéfices imprévus et des dividendes imprévus. Conformément aux prévisions, les auteurs constatent que la preuve de corroboration varie selon le degré d'imperfection de l'information relative aux bénéfices; les preuves de corroboration sont plus fortes lorsque les bénéfices varient davantage. Les auteurs font également état de constatations conformes à l'hypothèse selon laquelle les ana lystes écartent l'information relative aux bénéfices pour y substituer l'information relative aux dividendes dans le cas d'entreprises dont l'information relative aux bénéfices est imparfaite (variance élevée). Dans l'ensemble, les résultats invitent à la conclusion que les analystes utilisent l'information relative aux bénéfices et aux dividendes de manière interdépendante, une partie de cette interdépendance étant déterminée par l'imperfection de l'information communiquée en ce qui a trait aux bénéfices.

Purchases and sales of operating assets by firms generated $162 billion for shareholders over the past 20 years. This contrasts sharply with the evidence on mergers. This paper characterizes the behavior of value-maximizing firms, which could grow organically, purchase existing assets, or sell assets. The approach yields an endogenous selection model that links asset purchases and sales to fundamental properties of the firm. Empirical tests confirm the predictions of the model. In particular, return on assets and size strongly predict when firms purchase or sell assets, and the transaction size covaries with the value of capital employed by the firm. These findings indicate that corporate asset purchases and sales are consistent with efficient investment decisions.