Knowledge that Transforms

An efficient matching model of quits and layoffs is developed to account for several empirical regularities. Differences between quits and layoffs over the life and business cycles and across demographic groups are generated by differential rates of general productivity growth. The standard approach to quits and layoffs, based on wage rigidity, is shown to be incapable of accounting for many of the empirical regularities. Although a formal test rejects a structural prediction of the efficient turnover model, the specification does well in predicting both the level of and time-series variation in the fraction of separations labeled quits.

The unionization rate in a worker's industry and the worker's tenure are interacted with employer size in wage regressions using Current Population Survey data. The results provide insight into the relatively small size-wage relationship in the union sector. Nonunion tenure profiles steepen sharply with establishment size, while union profiles flatten. Also, the effect of industry unionization increases with firm size in the nonunion sector, a pattern indicative of threat effects. A separate analysis of nonunion wages indicates that size-wage relationships are stronger in occupations associated with high monitoring costs.

This article explores the McDonald and Solow (1981) conjecture that the outcome from a bargain over wages and employment can be approximately achieved by a bargain over wages and the labor-capital ratio (the "crew size"), where the latter can be argued to be more representative of the real world. I derive a crew size contract curve and show that this conjecture is not generally true. A test to determine which model is generating real-world data is suggested. The wage rigidity results derived by McDonald and Solow for their contract curve do not apply to this crew size contract curve.

This article examines the real wage-employment relationship in the United States in the light of the diversity of results in the literature. It demonstrates that, when the relationship is correctly specified, in the sense that due allowance is made for technical progress and capital accumulation, and the appropriate value-added deflator is used for the wage, then a negative relationship may be observed in the data. An incorrect specification, however, is as likely to lead to a positive relationship as a negative one.

Hazard-function estimates are utilized to analyze the effect of numerous public policy variables on strike duration, based on 7,546 strikes in Canada between 1967 and 1985. We find that only the mandatory strike vote substantially reduced conditional duration. However, the policy variables generally had a more favorable effect on reducing strike incidence, so that on net they tended to reduce unconditional duration (incidence times conditional duration). Specifically, reduced unconditional duration was associated with the existence of conciliation and the requirement of a mandatory strike vote, but the prohibition on replacement workers had a perverse effect.

One method of estimating losses resulting from work stoppages is to multiply the total number of man-days lost during disputes by the average product of workers. This statistic can be easily calculated using information typically available from government agencies but has obvious flaws. For example, the behavior of other firms not involved in disputes is ignored. Moreover, when output is durable, the impact on consumption is unknown since inventories can be used as buffers. To assess the importance of these and other considerations, I develop an alternative empirical framework and implement it using data from the British Columbian lumber industry.

We introduce borrowing constraints into the life-cycle theory of labor supply and show that they account for observed profiles in consumption, earnings, and hours worked. They can also account for differences in occupational choice across individuals who differ in initial wealth, marital status, or ability. This formalizes several aspects of observed differences in the labor-force behavior of men and women.

The 1980 census reveals a serious lag in the employment performance of young black men relative to young white men. With census data we test for demand-side causes of this lag, using both aggregate data for 94 standard metropolitan statistical areas (SMSAs) and disaggregate (or individual) data from the 1-in-100 Public Use Sample. Variation across SMSAs in the employment and wages of white youth provides indicators of the demand conditions for black youth, and we estimate that feasible increases in these demand factors would lead to about a 25% increase in the employment of black youth.

This article is concerned with hours worked per employee in unionized labor markets. First the determination of hours is examined in the context of various bargaining models and, in the process, these models are nested in a general framework. Then cross-section and time-series data are drawn on to quantify the effects of unionism on hours worked. The time-series data from 1920 to 1980 imply a negative impact of unionism on full-time hours while cross-section data for 1978 suggest some notable differences in both the direction and the magnitude of this impact across occupations and industries.

There is much variation in the union/nonunion wage gap across groups of workers within each of the two sectors, public and private. Furthermore, the variation in the public sector does not parallel in all of its detail that in the private sector. Thus, though the public-sector gaps typically are somewhat below their private-sector counterparts, there are important exceptions to this difference, especially among employees of local governments: public school teachers, clerical workers, refuse collectors, local transit bus drivers, licensed practical nurses, hospital technicians, nonprofessional hospital workers, and undoubtedly some others.