Gross investment packs

The regression results indicate that FDI has a statistically significant and positive impact on wages that is not significantly affected by the model’s other variables. Looking at the full model with gross investment (regression V), a one percent increase in FDI increases wages by about 0. 02 percent. Put another way, if FDI increases by one standard deviation from its mean, wages will increase by 1. 32 percent. In regression VA, the coefficient increases slightly so that a one percent increase in FDI increases wages by about . 025 percent.

The actual effect probably lies somewhere between these two figures, as some small proportion of FDI is comprised of financial flows that are not counted as a part of gross investment. Gross investment packs a lot more punch: in regression V a one percent increase in investment raises wages by 0. 19 percent - an investment increase of one standard deviation from the mean raises wages by about 22 percent.

Looking at the adjusted investment measure in VA lowers this estimate: a one percent increase in domestic investment raises wages by 0.16 percent. But it is still a much larger impact than FDI. Looking at the rest of the variables, we will confine the discussion to the regressions that use adjusted investment, keeping in mind that the true results are somewhere in between the regressions that use the gross and adjusted investment measures.

Trade has increased the overall demand for labor in China, thereby raising wages: in regression VA, a one percent increase in trade raises wages by . 036 percent. Put in terms of standard deviations, an increase of one standard deviation raises wages by 2. 9 percent. The labor force variable, reflecting a reserve army effect where greater supplies of labor drive down wages, has the expected effect, and it is notably sizeable, suggesting that creating ample employment for China’s growing population is a validly key concern for policy makers. Productivity has the correct sign but is not statistically significant, probably due to noise in the measure. The results on liberalization are interesting for a couple of reasons.

First, including liberalization does not affect the other variables that one might posit are picking up liberalization effects: neither FDI nor trade, sometimes treated as proxies for liberalization, are altered significantly by the inclusion of the liberalization variable. At least for the Chinese case, then, these openness measures are not picking up the effects of liberalization on the Chinese economy. Secondly, there is a consistently positive and statistically significant relationship between liberalization and wages.

If the proportion of state sector output to all industrial output decreases by one percent, wages will increase by 0. 086 percent; if L2 declines by one standard deviation from the mean, wages will increase by 3. 06 percent. This result suggests that the freedom from wage controls, and the overall increase in incomes (especially rural) that came with liberalization, were important positive benefits of the economic reform program begun in 1979.

But it is important to recall that we are only measuring money wages. The substantial non-cash benefits that can come along with a state sector job, such as housing allowances, still render these jobs among the most prized, at least for less educated or lower-skilled workers who would otherwise have trouble accessing workplace benefits. Some interesting work on the impact of FDI on relative wages in China illustrates this point. In a study that compares state and FIE sector wages for unskilled workers, it was found that after including non-wage benefits like pensions, housing and medical care, state sector jobs were much better for unskilled workers who had been assigned or expected to receive public housing (Zhao and Xu 2000).

The same estimate was not done for the college-educated, but it was hypothesized that wages between the state and foreign sectors for skilled workers would be much smaller since FIEs offer good benefits packages to the college-educated. The work of Yaohui Zhao (2000) bears on this issue. He argues that because of the segmented nature of labor markets in China, FDI does indeed raise the relative wages of skilled workers, but not because FIEs demand more skilled workers. Using urban household survey data in 1996, Zhao finds that education is used to access the privileged state sector.

Unskilled workers are thus more abundant in the unprivileged or informal sectors, giving FIEs easy access to unskilled workers. Conversely, FIEs must compete with the state sector for skilled workers, bidding up the relative wages of skilled workers in the foreign sector. He also finds that the proportion of skilled to unskilled workers is similar in the state and foreign sectors, suggesting that FIEs do not employ more skilled labor than state firms, casting doubt on the notion that relative wage increases are a result of FIEs enjoying higher productivity.

While FIEs do pay higher wages (including bonuses and cash subsidies) than state firms, these higher wages are unevenly distributed among workers of different educational levels. Including noncash benefits such as housing, less educated workers earn less in FIEs than in state enterprises, but more educated workers earn more. Summary In sum, then, we do find empirical evidence that over the last decade and a half, FDI has raised wages in China, but the impact is small in relation to the effects of domestic investment and to a lesser extent liberalization.

The impact of foreign trade on wages is also significant, both statistically and economically. In fact, this latter variable might be picking up some of the impact of FDI on wages, since, as we discussed above, a great deal of FDI was directed toward the export sector. Even if one adds up the effects of these two variables, implicitly attributing all the export effect to FDI, the sum of the coefficients on FDI and trade is about 0. 06, which is about 1/3 of the impact of investment on wages (0. 16). Of course, this is an overestimate of the FDI effect because not all trade is due to FDI and foreign invested enterprises (though a significant amount is, perhaps as much as 60 percent).