Sunday, October 11, 2009: 11:35 AM
In this paper, we investigate how quickly firms react to various shocks by changing wages, and how the speed of wage adjustment depends on the type and direction of the underlying shocks. Our analysis draws on two surveys of Estonian company managers that we carried out in 2005 and 2007. The surveys inquired about wage adjustments in response to four types of shocks: exogenous changes in sales turnover, output price, labor productivity, and competitors’ wage level. Company managers were asked whether changes in each of these variables would have an effect on wages in their companies and if so, how soon wages (including paid bonuses) would be adjusted. Importantly, the surveys differentiated between the speed of wage adjustment in response to increases and decreases in the above four variables. To our knowledge, there exist no previous surveys that have directly asked businesses about the time lags in wage adjustment to specific economic shocks.
The answers indicate that firms’ reactions to shocks tend to be asymmetric. In particular, the speed of response is related to the direction of a shock – a number of firms react faster to a negative shock than to a positive shock. The nature of the shock is considered from a company’s viewpoint. For example, an increase in turnover is a positive shock, whereas an increase in competitors’ wage level is a negative one. We test the statistical significance of differences in the speed of wage adjustment to positive and negative shocks using t-tests and find that they are highly significant (at 99% level) for all types of shocks covered by our surveys.
After describing the presence of asymmetry in the speed of wage adjustment, we investigate which firms are more likely to demonstrate it in their behavior. Using four binary indicators for asymmetric reaction as dependent variables, we employ probit and heckman probit models to relate the incidence of asymmetry to a number of firm-specific characteristics available in the survey data. The estimation results imply that firms operating in a more competitive environment have a higher propensity to react asymmetrically. We also find that the asymmetric reaction is influenced by production technology: firms relying on labor-intensive technology are more likely to react faster to a negative shock than to a positive shock. Firm size (measured on the basis of employment) is negatively related to the propensity of an asymmetric reaction. The pattern of estimated effects is similar across different types of shocks.
The answers indicate that firms’ reactions to shocks tend to be asymmetric. In particular, the speed of response is related to the direction of a shock – a number of firms react faster to a negative shock than to a positive shock. The nature of the shock is considered from a company’s viewpoint. For example, an increase in turnover is a positive shock, whereas an increase in competitors’ wage level is a negative one. We test the statistical significance of differences in the speed of wage adjustment to positive and negative shocks using t-tests and find that they are highly significant (at 99% level) for all types of shocks covered by our surveys.
After describing the presence of asymmetry in the speed of wage adjustment, we investigate which firms are more likely to demonstrate it in their behavior. Using four binary indicators for asymmetric reaction as dependent variables, we employ probit and heckman probit models to relate the incidence of asymmetry to a number of firm-specific characteristics available in the survey data. The estimation results imply that firms operating in a more competitive environment have a higher propensity to react asymmetrically. We also find that the asymmetric reaction is influenced by production technology: firms relying on labor-intensive technology are more likely to react faster to a negative shock than to a positive shock. Firm size (measured on the basis of employment) is negatively related to the propensity of an asymmetric reaction. The pattern of estimated effects is similar across different types of shocks.