Management quality matters: Evidence from labour cost shocks in China

Harald Hau, Yi Huang, Gewei Wang

17 October 2016



There is no shortage of management books advocating a creative management idea which, they all claim, will provide a firm with the competitive edge it needs to thrive in an ever-changing business environment. These books make use of carefully selected case studies and examples to support their ideas. Generally, though, the authors do not clearly identify a common competitive shock nor compare the firm-specific response using a statistically meaningful sample. Paul Krugman referred to these books as ‘airport economics’.

Studying the way that corporations adjust to competitive shocks is difficult, and isolating the role of management even more so. New technology directly affects the productivity of firms, but it simultaneously transforms the competitive environment to which management needs to adapt. It is difficult to disentangle the role of external competitive forces from the direct effects of technological change. Policy shocks provide better opportunities to isolate the firm response because they do not directly emanate from an underlying process that is shaping the productivity of a firm. Recent research has therefore focused on policy shocks, such as a new trade agreement, unexpected exchange-rate shocks, or changes to the minimum wage.

On 15 January 2015, the Swiss National Bank abandoned its peg for the Swiss franc, which then appreciated. This is an example of an unexpected, external competitive shock. It raised the production costs of Swiss firms relative to their pricing power in export markets, and this forced them to operate in a more difficult competitive setting.

Are exporting firms able to squeeze costs further and raise firm productivity in response? Should we expect management quality to play a role? Unfortunately for economists, management quality is unlikely to differ much across exporting firms in a small country like Switzerland. Furthermore, badly managed Swiss firms might simply not have survived previous Swiss franc appreciations, and therefore don’t exist in the sample. Ironically, the more management quality matters for firm survival, the harder it is to detect its relevance in the increasingly homogeneous sample of surviving firms.

Minimum wage shocks in China

Emerging markets might provide a better setting for a test of the relevance of management practice, because these markets have firms which differ more widely in their governance and management quality. Our recent research focuses on the productivity response of Chinese firms to external labour cost shocks (Hau et al. 2016). Many industries in China feature a large numbers of state-owned firms (SOEs), private Chinese firms, and foreign firms. Each have very different management practices. For example, generous state-guaranteed credit arguably keeps even poorly managed SOEs in business.

Chinese firms have undergone regular large labour cost shocks because of large revisions to the minimum wage set at the local level. In the seven-year period from 2002 to 2008 alone, there were more than 17,000 minimum wage increases in China’s 2,852 counties and cities. Figure 1 shows the distribution of these minimum wage changes for the period 1996 to 2012. We distinguish between counties and cities which had their minimum wage increased by between 0% and 10%, between 10% and 20%, and by more than 20%. Minimum wage increases in excess of 20% were also frequent.

For each year in Figure 1, we plot the distribution of minimum wage increases in Chinese counties and cities. Minimum wage increases between 0% and 10% are light blue, between 10% and 20% are violet, and increases of more than 20% are red.

Figure 1 The distribution of minimum wage increases in China, 1996-2012

For firms with many workers at, or near, the minimum wage, these minimum wage increases often had a dramatic impact on average labour cost. We estimated the impact of the minimum wage increase on the change of the average firm wage as a function of the initial average firm wage for a large sample of industrial firms between 2002 and 2008. Earlier years were excluded because of a more lenient enforcement of minimum wage regulation. The firm survey was discontinued in 2009. There are 1,192,144 firm-year observations for this period.

Figure 2 shows the increase in average labour costs (by firm size) for a 20% increase in the minimum wage, and also plots the distribution of the initial average firm wage relative to initial minimum wage. On the horizontal axis a value of 1 represents a firm for which the average firm wage is the minimum wage, and a value of 6 represents a high-wage firm with an average initial wage six times the average wage. Small firms are those with fewer than 100 employees, and large firms are those with more than 1,000 employees.

Figure 2 The effect on the average wage of a 20% increase in the minimum wage on Chinese firms, 2002-2008

In the figure we estimate the effect of a 20% minimum wage increase of the firm’s average wage as the function of the ratio ws/wmin of the average firm wage to the initial minimum wage. This is the red solid line, with the 95% confidence interval as dashed line. The blue bars are the distribution of firms over the ratio ws/wmin in each category.

It is no surprise that for firms for which the average wage is also the minimum wage, a 20% increase in the minimum wage translates directly into a 20% increase in the firm’s average labour costs. This minimum wage effect convexly decreases for firms with higher wages, to the right of the firm wage distribution. The cost shock does not differ much by firm size, but is primarily a convex function of the average wage relative to the initial wage. We refer to this as the impact function IF(ws/wmin) of minimum wage increases.

Firm productivity and its response to competitive shocks

Management theorists have argued for many years that firms often operate far from the efficient frontier (Leibenstein 1966). Better productivity measurement in the last 20 years confirms this view. The dispersion of plant-level productivity in developed markets differs by a factor of two between the 10% and 90% quantiles of the productivity distribution (Syverson 2011). In emerging markets, it differs by a factor of five. The larger the market and financial distortion, the greater the heterogeneity in terms of productivity, and also management quality.

In the long run, however, the productivity distribution depends on the dynamic firm response to competitive shocks like these minimum wage shocks. Do the most-affected Chinese firms increase their productivity, and what causes this response?

To investigate this question, it is instructive to split the firm sample by ownership and productivity. We split the sample into SOEs, private-owned Chinese companies and foreign-owned companies. We also split them depending on whether they were above (high-TFP) or below (low-TFP) average industry productivity. We used panel regressions to estimate the average productivity response to the minimum wage increase. The estimation accounts for the heterogeneous firm exposure to the minimum wage which we estimated in Figure 2, which allowed us to discriminate between firms with different wage structures that have been subjected to the same minimum wage shock.

Figure 3 shows the magnitude of the productivity increase (measured relative to industry peers) achieved by the most exposed firms in response to a 20% minimum wage increase. The incremental productivity increase of SOEs was small and statistically insignificant (indicated by the vertical line representing a 95% confidence interval around the mean effect). Private firms show a statistically significant productivity response if their initial productivity is low, whereas foreign firms show by far the largest productivity increase in the year of the minimum wage increase. The point estimate of incremental TFP growth, just under 10%, is economically large and corresponds to almost a full year of trend growth. External competitive shocks therefore trigger large productivity improvements in some firms, but not in others. Ownership structure accounts for a large proportion of this variation.

Figure 3 Estimated incremental productivity increase in response to a minimum wage increase by 20% for the most exposed Chinese firms, 2002-2008

These results do not depend much on how we measure firm productivity. When we examined related variables like output growth, investment behaviour, or export volumes, we got similar differential results. Chinese export statistics are detailed, and that allowed us to construct a strictly volume-based export measure without any price distortions. A similar pattern of much stronger export response to labour cost shocks emerges for private and, in particular, foreign firms. Could differences in management practice account for the heterogeneous firm response in Figure 3?

Management quality and its measurement

Bloom and Van Reenen (2007) measured the quality of firm management by three dimensions:

  1. Monitoring practices: The collection and processing of production information;
  2. Target-setting practices: The ability to set coherent, binding, short-term and long-term targets; and
  3. Incentive practices: Merit-based pay, promotion, hiring, and firing.

These three measures can be aggregated to a combined score of overall management quality. Later work extended this scoring method to firms in emerging markets, including China (Bloom and Van Reenen 2010). Their Chinese firm sample overlaps with our productivity data for 564 Chinese firms sampled in 2006, 2007, 2008, and 2010. We found that larger firms, and particularly foreign-owned firms, had high average management scores. SOEs have lower average management quality.

We regressed the observed management score for the 564 firms onto characteristics like ownership type and firm size. Then we extrapolated from these determinants of management quality to the full sample, to create a variable that measures predicted management score. Finally, we interacted firm response to minimum wage shocks with the predicted management score, and repeated the previous panel regressions.

Figure 4 shows the firms’ incremental TFP growth as a function of the triple interaction of the firm exposure to the minimum wage shock, the minimum wage increase, and the management score.  

Figure 4 The residual plot shows the firm productivity response for 29,998 firm events in which a Chinese firm experiences a more than 20% increase in the minimum wage, 2002-2008

The horizontal axis shows the triple interaction between the impact function IF(ws/wmin), the (log) minimum wage increase and the predicted management score (Mgmt_Score). To keep the figure simple, we plotted only those 29,998 firm-year observations for which the minimum wage increase was larger than 20%. For the same minimum wage exposure (or value of the impact function) and the same large minimum wage increase, a firm observation with a low predicted management score will tend to be on the left. Those with a high predicted management score will be to the right. The positive slope coefficient in Figure 4 illustrates that the positive productivity response to the labour cost shock tends to be, ceteris paribus, larger for firms with a higher predicted management quality. The slope coefficient has been estimated for the full sample of firm-year observations, where all single and double interaction terms are also included.

This evidence indicates better-managed firms adapt better to adverse competitive shocks, and suggests that management quality matters for this adaptability. It is difficult to reconcile the heterogeneous response of Chinese firms with so-called efficiency wage theories. These assign the productivity increase directly to the motivational effects of higher minimum wages. Such motivational effects should operate independently of management practice, or firm governance.

Private and foreign firms may differ in unobservable dimensions which increase their adaptability compared to SOEs. Yet, in this case, external factors and constraints mostly disadvantage private and foreign firms. Therefore, we believe that differences in the internal management practice represent a plausible explanation for the highly heterogeneous productivity response among Chinese firms. So, there’s good news for authors of airport economics books: Chinese firm evidence supports your claim that management practice matters.


Bloom, N., and J. Van Reenen (2007), “Measuring and Explaining Management Practices across Firms and Countries”, Quarterly Journal of Economics, 122(4), 1351-408.

Bloom, N. and J. Van Reenen (2010), “Why do Management Practices Differ across Firms and Countries?”, Journal of Economic Perspective, 24(1), 203-24.

Wang, G., H. Hau, and Y. Huang (2016), “Firm Response to Competitive Shocks: Evidence from China’s Minimum Wage Policy”, SFI research paper no. 16-47.

Leibenstein, H. (1966), “Allocative Efficiency vs. ‘X-Efficiency’”, American Economic Review, 56 (3), 392-415.

Syverson, C. (2011), “What Determines Productivity?”, Journal of Economic Literature, 49(2), 326-65.



Topics:  Productivity and Innovation

Tags:  competitive shocks, Management, productivity, minimum wage, efficiency wage

Professor of Economics and Finance, Swiss Finance Institute, University of Geneva; and Research Fellow, CEPR

Assistant Professor of International Economics and Pictet Chair in Finance and Development, The Graduate Institute, Geneva

Research Assistant Professor, IGEF