Faster real wage growth associated with faster productivity growth across countries: Panel regression results across 21 countries and 35 years
| Linear combinations | (1) | (2) | (3) | (4) | |
|---|---|---|---|---|---|
| (1) | Four real wage growth lags | 0.12** | 0.14** | 0.27*** | 0.23*** |
| (.057) | (.058) | (.06) | (.075) | ||
| (2) | Four productivity growth lags | 0.44*** | 0.43*** | 0.03 | 0.09 |
| (.077) | (.07) | (.08) | (.056) | ||
| (3) | Four GDP growth lags | 0.11** | 0.11** | 0.36*** | 0.32*** |
| (.044) | (.044) | (.06) | (.058) | ||
| (4) | Implied long-run effect of wage growth (line 1/line 2) | 0.21 | 0.25 | 0.28 | 0.25 |
| Gap | no | yes | yes | yes | |
| Country fixed-effects | yes | yes | yes | yes | |
| Year fixed-effects | yes | yes | yes | yes | |
| Country-specific time-trend | no | no | yes | yes | |
| Panel-specific heteroskedasticity | no | no | no | yes | |
| Countries | 21 | 21 | 21 | 21 | |
| Years | 37 | 37 | 37 | 37 | |
| Observations | 386 | 386 | 386 | 386 | |
| Wald chi2 | 313.63 | 323.62 | 482.86 | 477.86 |
Note: * = significant at 10 percent level, ** = significant at 5 percent level, *** = significant at 1 percent level.
Dependent variable is annual productivity growth. Controls, as described in table, include lagged values of dependent variable plus lagged values of real wage growth and GDP growth. Lagged values of real wage growth are used to avoid the possible endogeneity of productivity and wage growth. Lag lengths followed the methods of Vergeer and Kleinknecht (2010) and are chosen to avoid correlation in the residuals. Tests of correlation for the lag lengths used in this paper fail to reject the null hypothesis of no correlation.
Source: Data from the Conference Board Total Economy Productivity (TEP) database.