مشابه عین حال متفاوت : گزارشات کارگری در شیوه کار و نتایج را در یک شرکت در سراسر کشور

This paper examines cross-country differences in labour policies and practices and employee performance and attitudes toward work from a sample of nearly 30,000 employees in a large multinational manufacturing firm. The analysis shows: 1) large establishment and country differences in work practices, performance, and attitudes toward work across countries; 2) qualitatively similar responses of workers to work practices across countries; 3) a strong link between the establishment average of employee reports on the quality of labour-management relations and establishment average measures of employee performance 4) a positive relation between average employee performance and average employee-management relations at the country level, but no relation between country level performance in the firm and measures of the extent of national labour regulations or practices.

Consider a multinational firm producing similar goods and services in the same industry in many countries. How much do labour practices, employee attitudes, and worker behaviour differ among the establishments of the firm? Are there identifiable differences in practices, attitudes, and behaviour across the countries? Do workers in different countries respond similarly to similar policies? Are aggregated measures of labour practices by country related to aggregated measures of worker behaviour and outcomes by country? This paper examines these questions using data from a 2005–2006 survey of workers in 272 establishments in a single large multinational manufacturing firm that operates in 19 countries. With the assistance of the firm, we undertook a web based survey and a paper survey of the workers. The firm convened employee meetings in each facility to induce employees to respond to the survey, which gave us a sample of 29,353 respondents, with a response rate of greater than 60%.1 This is one of the largest individual level data sets on labour practices, employee attitudes toward work, and self-reported workplace performance across countries in a single firm since Geert Hofstede, 1984 and Hofstede, 1991 study of IBM based on surveys in 1968 and 1972.2 Our study differs from Hofstede's classic work in three ways. First, whereas Hofstede's surveys focused on European and Middle Eastern countries, our sample contains a large number of observations from the US and from developing and developed countries in Latin America and Asia as well as from Europe. Second, our data are for manufacturing, whereas Hofstede's 1968 study included many workers outside manufacturing and his 1972 sample excluded manufacturing. Finally, while Hofstede looked primarily at employee values and beliefs as they related to organizations and national culture or character, our focus is on employment relations, the organization of work, and the economic behaviour of workers in response to labour policies and practices in their establishment.3 Our data reveal that: 1) Workplace practices, worker reports of performance, and attitudes toward performance differ significantly across establishments and among establishments in the different countries in which the firm operates. In all of our analyses, estimated dummy variables for the establishment where a worker works show that the establishment is a major factor in responses to questions about workplace practices, attitudes toward work, and performance. Replacing establishment dummy variables with dummy variables for the country in which the establishment is located identifies significant cross-country variation in labour practices and outcomes as well. 2) Worker attitudes toward their workplace and workplace behaviour are affected by work policies and practices in qualitatively analogous ways across countries. Statistically, regressions linking measures of worker attitudes or performance to measures of the quality of labour-management relations and the presence of high performance work practices yield positive estimated slope coefficients in most countries, although with differing magnitudes. 3) Aggregating worker responses by establishments to make the establishment the unit of observation, we find that the average employee performance across establishments is strongly related to the average reported quality of labour-management relations across establishments. This relation holds for establishments outside the US and for establishments within the US. 4) Taking country as the unit of observation, workers in countries where workers report better employee-management relations and compensation above market levels also report better employee performance. By contrast, we find no relation between the average measures of worker performance among countries with widely used country level indices of labour practices.

This study has examined work practices, employee performance, and attitudes toward work at the level of individual workers, establishments, and countries using a survey of nearly 30,000 employees in a large US multinational manufacturing firm that operates 272 establishments in 19 countries. The survey is one of the largest surveys of workers across countries in a single firm. Our analysis found that employment policies and practices, and employee performance and attitudes toward work, vary among establishments and across countries. To help identify any independent effect of country labour institutions or practices on workers and establishments, we compared the variation in practices and outcomes across the countries in which the firm operates with the variation in practices and outcomes across states or groups of states in the US in which the firm operates. This comparison revealed greater variation across countries than across states - a pattern consistent with the notion that country institutions have an independent effect on labour practices and performance, beyond any effects related to contiguity of facilities. We also found that worker performance and attitudes were strongly related to policies and practices in broadly similar ways across these countries, though we noted some divergences for particular policies or outcomes. To deal with the problem that patterns found among individuals need not generalize to establishments or countries, we analysed the relation among establishment level averages of variables. This analysis found a strong relation between the quality of labour-management relations and employee behaviour and outcomes consistent with our analysis of data for individuals. But the analysis of establishment averages also found weak or negligible relations between high performance work places and employee behaviour and outcomes, whereas the analysis of individuals found stronger relations. This could reflect genuine variation in workplace policies or practices within establishments, or idiosyncratic reports on policies, practices, and performance by individuals that biases upward the estimates of the relation between policies/practices and outcomes at the level at which those policies and practices are set. We found only weak relations in both the individual and establishment analyses between the worker reporting above-market compensation and performance or attitude toward work. At the country level, our analysis revealed a positive relation between the average quality of employee-management relations at the country level and average employee performance. By contrast, we found no relation between our measures of employee behaviour or attitudes and measures of national labour relations that differentiate countries by their degree of institutional vs. market determination of outcomes. This suggests that the establishment level policies and practices of the multinational trump the effect of these broader features of labour markets in determining how workers behave within the firm. Overall, our analyses support the idea that while labour practices and performance vary across countries, good labour management relations is an important factor in worker behaviour and workplace performance in virtually all situations, while the effects of high performance practices and compensation on performance and attitudes are weaker. Moreover, employee-management relations has a strong relation not only to standard performance measures such as likely turnover and willingness to work hard, but also to our innovative measure of how workers would respond to a shirking co-worker. Further research is needed to identify the specific ways in which management treats workers and communicates with them that underlies worker feelings that labour-management relations are good.