Enterprise Performance

Despite the evidence on improving profitability and the trends towards privatization from Table 1 [ PDF 87.4KB | 1 page ]and Table 2 [ PDF 55.9KB | 1 page ]there is a widespread perception examined in the early section of this paper that the north-east is a problem region. To resolve this apparent discrepancy and to cast light on the investment prospects for the region we turn to a detailed examination of enterprise performance. We use two simple performance indicators value-added per employee (VA/L), as a measure of productivity, and 'imputed profits' to total assets (IP/TA), as a measure of returns on investment.¹¹ Imputed profits are calculated as value added minus the sum of wages, financial charges and depreciation. We prefer this to enterprises' own accounting profits from their published accounts (which are the basis for Table 1 [ PDF 87.4KB | 1 page ]) as the frequent changes in accounting practices can distort the underlying picture.

Indicators such as these must be related to particular industrial sectors and a direct comparison of these two measures for the north-east and the rest of the country is given in Table 3 [ PDF 81KB | 2 pages ]and Table 4 [ PDF 78KB | 2 pages ]. These tables give the average value for the two indicators in 37 two-digit sectors for the region and the rest of the country. In only three sectors (Food production, Petroleum processing and Non-Metal mining) does the north-east have a higher average productivity. Similarly it has a higher average profitability again in only 3 sectors (now Petroleum extraction, Petroleum processing and Furniture). In all other cases performance of firms in the northeast is below that in the rest of the country, indicating a clear 'performance gap'. Hence despite the improvements in profitability implied in Table 1 [ PDF 87.4KB | 1 page ], relative to the rest of the country the region is not doing well when the relevant comparison is made across sectors.

The key issue we wish to address is what accounts for this gap. How far is it due to factors like differences in scale, technology, ownership, and competition across regions, and how far is it due to the operating environment or 'investment climate' which create a 'location disadvantage' in the north-east? To address this we employ a fixed effects panel data regression model to the enterprise data over 1995-2002.

This can be thought of as a simple 'structure-conduct- performance' approach that attempts to isolate the different effects on performance, based on the characteristics of the enterprise itself, the characteristics of the sector in which it operates, time factors and a series of dummy variables including province specific and regional dummies. As our regression results are based on a very large and representative data set with over 44,000 firms, and on a rigorous econometric approach , the conclusion here should be much more robust and systematic than that drawn from casual observation or limited case studies.

Generically the model can be written as

P_it = á₀ + á₁X_it + á₂Y_jt + á₃Z + í_it

where P is a performance indicator (productivity or profitability)

X is a vector of firm-specific factors for firm i (relating to scale of production and factor intensity)

Y is a vector of sector-specific factors for sector j (relating to concentration and ownership)

Z is a vector of dummy variables (relating to provinces, sectors, years and interaction terms)

á0 is a constant, í is an error term and t indicates annual observation.

Four different versions of (1) are applied (models 1 to 4). The relevant variables used are set out below:

Dependent variables

(VA/L): value–added per worker:
(IP/TA): imputed profit to total assets:

Independent variables

• ln(L): size of the firm (as measured by number of workers);
• ln(Kp/L): intensity of production fixed capital (total capital assets per worker)
• ln(Kf/L): intensity of non-production fixed capital (welfare capital per worker)
• Ind3Concentration: Herfindal index for industry concentration at 3-digit industry level calculated for the sample
• FIE_ind2MKT_Share: market share of foreign invested enterprises at 2 digit industry level in the sample
• D^ind2 : Sector dummies at 2 digit industry level
• (D^ind2) * ln(Kp/L): Interaction terms between sector dummies and capital intensity
• D^type: Ownership dummies for SOEs, private, collective, mixed, foreign and Hong Kong, China,Taipei,China ownership.
• D^year: Annual dummies 1995 to 2002
• D^nep: North-east region dummy
• D^nep * ln(Kf/L) interaction term between north-east dummy and intensity of nonproduction capital
• D^nep * D^ind2 interaction term between north-east dummy and sector dummy
• D^nep * D^type interaction term between north-east dummy and ownership type dummy
• D^nep*D^year interaction term between north-east dummy and annual dummy
• D^Lioaning, D^Jilin and D^Heilonjiang : province specific dummies for the north east
• D^place2 : province specific dummies

The regression equation for model 1 is

The main purpose of this regression model is to identify the performance gap between the enterprises in the north-east and in the rest of PRC after controlling for other factors that are not specific to location effects. In other words, the regression coefficient for the north-east region dummy (D^nep) indicates the performance gap for enterprises in the north-east that is specifically due to the location effects after controlling for enterprise scale and factor intensity, sector competition, ownership and other unmeasurable sector characteristics and time. General macro economic effects are captured through the time dummies.

Scale can have an ambiguous impact on performance depending on the effect of economies of scale. Regarding factor intensity insofar as capital intensity reflects a higher level of technology higher production capital per worker can be expected to have a positive impact. Social obligations of enterprises or 'welfare capital' per worker may raise productivity but may also lower profitability and there is the possibility of a negative impact on the profit measure. Sector concentration is taken as a proxy for the degree of competition so that high concentration implies low competition; it is hypothesized that competition has a positive impact on productivity but it may have a negative impact on profitability. Similarly the foreign ownership share in a sector is expected to have a positive impact on productivity both through competition and potential technological spillovers. Its impact on profitability is more ambiguous. Model 1 also includes controls for type of ownership of a firm, annual effects, and sector characteristics. The reference points with which the dummies are compared are SOEs for ownership, 1996 for years, textiles (sector 17) for sectors and the rest of PRC for the north-east dummy.

The results of model 1 are shown in Table 5 [ PDF 77.2KB | 2 pages ]and the implied differences in performance relative to the various reference points are given in Table 6 [ PDF 76KB | 1 page ]. In terms of productivity the scale variable is negatively related to productivity, indicating there are no scale economies. Both capital intensity variables, both production and non-production capital, are positively related to productivity. The concentration ratio is negative related to productivity implying there is a productivity-enhancing competition impact from lower concentration and higher competition. The share of foreign firms in sector output is positively related to productivity, which is consistent with a positive technological spillover. All of the ownership dummies are positively related to productivity indicating that it is higher in enterprises that are not SOEs. The largest coefficient and thus the highest productivity gap over SOEs is for foreign firms, who have 80% higher value-added per worker relative to SOEs, controlling for other factors. In terms of time, productivity shows a significant improvement from 1999 onwards. The regional dummy for the north-east, which is our key variable of interest, is significantly negative, indicating that if all the other control variables have the same value, enterprises located in the northeast are likely to have lower value-added per worker than those in the rest of the country by as much as 44 percentage points.

A different pattern is observed for the profitability measure. Here there is a very weak scale effect as the coefficient on the size of firm is positive, but very small, and only significant at the 10% level. The implication is that as enterprise size grows there may be economies in the use of intermediates and capital rather than labour. However both measures of capital intensity are now negatively related to profitability. As in the case of productivity, concentration is negatively related to performance, so profitability falls with concentration, allowing for all other factors. This is an unexpected result and is likely to be due at least in part to the lower productivity associated with higher concentration noted above. Foreign firm share in a sector is positively related to profitability. The coefficients on all ownership dummies are positive indicating that all ownership types are more profitable than SOEs, when controlling for all other measurable features. Now it is private domestic firms rather than foreign firms that show the largest gap relative to SOEs. Private firms are over 8 percentage points more profitable than SOEs controlling for all other measurable factors. In terms of time profitability shows a significant improvement one year later than productivity in 2000. The regional dummy for the north-east is again strongly significant and negative. It shows that controlling for all other factors the profit to assets ratio in the north-east is 4 percentage points below that of comparable enterprises in other locations in PRC. This and the productivity performance gap are specific to the north-east region and cannot be explained by the other controlling variables in the regression.

These results imply that standard reforms such as privatization, the introduction of market competition and foreign direct investment can help the north-east region to improve enterprise performance, but even allowing for its level of these factors, the region still has a performance gap that is specific to its own location.

Thus far we have treated the north-east as a single region. Model 2 addresses this by replacing the single north-east dummy with three separate provincial dummies (D^Lioaning, D^Jilin and D^Heilonjiang.)

This regression is designed to check if the three provinces in the north-east region have performed differently relative to the rest of the country. The results, which are not reported here, show an almost identical pattern to those in table 5. The provinces performed similarly and have almost the same performance gap with the rest of PRC. Liaoning has the largest gap. Its profitability is 4.7 percentage points lower, compared with 3.3 percentage points for Jilin and Heilonjiang. In general however there are no significant 'within north-east' effects with all three provinces sharing broadly similar locational disadvantages.

Model 3 controls for the same scale, technology, competition and ownership variables used in model 1, whilst in addition removing the negative location-specific impact of the north-east region. This is done by introducing a set of four interaction terms in addition to the original north-east dummy. These are

between the regional dummy and welfare capital intensity (D^nep and ln(Kf/L))
between the regional dummy and each sector dummy (D^nep and D^ind2)
between the regional dummy and each ownership type (D^nep and D^type)
and between the regional dummy and each annual dummy (D^nep and D^year).

The first set of interaction terms (D^nep * ln(Kf/L)) is designed to check for the effect of nonproductive or welfare capital on performance, that is specific to the north-east region. We often hear the claim that the enterprises in the north-east have larger social burdens, such as the provision of employee housing and fringe benefits like schools and hospitals. If this effect is important, it will show up as a statistically significant negative coefficient for this interaction term.

The second set of interaction terms is designed to compare the performance gap between the north-east region and the rest of PRC by sector, after already controlling for systematic factors, like scale and ownership, as well the standard (that is the non-interactive) north-east (D^nep) and sector dummies (D^ind2). Thus the coefficients on the interactive terms (D^nep * D^ind2) will indicate the performance gap after controlling for both industry sector and location effects. In another words, they will indicate the gap in productivity or profitability between enterprises in the north-east and those in the rest of the country, after the negative impact of the north-east region's location effect is removed. Regression analysis allows us to do this counter-factual exercise by decomposing the impact on performance arising from different sources. We can then compare enterprise performance in the north-east after locational disadvantages are removed with performance in the rest of the country to get a version of 'potential comparative advantage' by sector for the region.

The third set of interaction terms (D^nep * D^type) is designed to check if the non-state enterprises in the north-east region are doing exceptionally better or worse relative to SOEs than the non-state enterprises in the rest of the country after controlling for all other factors, including the location effect of the region. There is a view that the north-east region does not offer a supportive environment for the development of non-state enterprises, even if its current locational disadvantages could be removed, and a negative coefficient on this term will provide support for this.

The fourth set of the interaction terms (D^nep * D^year) is designed to examine the timing of the performance change of enterprises in the north-east region to see it differs from that nationally, after controlling for locational disadvantages.

The pattern of results, which are not reported here for reasons of space, is identical to that reported in Table 5 [ PDF 77.2KB | 2 pages ]for model 1. The interest here is in the impact of the new interaction terms.

For the interaction of the north-east dummy and the intensity of welfare capital there is a positive but weak impact (significant at 10% level) on productivity but no significant impact on profitability. Hence over and above the direct negative impact of this form of welfare expenditure on profitability, which is captured by the coefficient on (Kf/L) there is no specific north-east region dimension to this impact on profits.

The coefficients on the interactive terms (Dnep * Dind2) indicate the performance gap for the region after controlling for sector, location and all other measurable effects. Out of 37 sectors there are positive coefficients for 31 for the productivity and 26 for the profitability measures; of these 24 and 12, respectively, are significant. A positive coefficient on (D^nep * D^ind2) indicates a superior performance relative to the rest of the country, if locational and other disadvantages captured by the explanatory variables are controlled for. These figures must be seen in the light of the direct comparison in Table 3 [ PDF 81KB | 2 pages ]and Table 4 [ PDF 78KB | 2 pages ]on the performance by sector in the region as compared with the rest of the country. By the profitability measure in the direct comparison in only 3 out 37 sectors (Petroleum Extraction and Processing and Furniture) were enterprises in the north-east performing better than those in the rest of PRC. The results in Table 8 [ PDF 83.9KB | 1 page ]and Table 9 [ PDF 52.9KB | 1 page ]show that if we control for the negative location effect of the north-east and the level of other variables, in 26 sectors the north-east has the potential to be competitive in terms of an above average level of profits. The 12 sectors where there the positive coefficient is significant include the three noted above (Petroleum Extraction and Processing and Furniture) plus Food production, Nonmetal mining, Beverages, Medical supplies, Instruments, Nonmetal products, Transport Equipment and Electronics and Telecoms. These are a mixture of resource-based and high technology activities. The implication is that the poor performance of the north-east region in the direct sector comparison is largely due to a range of systematic factors, for example relating to ownership and competition, as well as to location-specific effects . If the north-east region can catch up in these location and non-location specific areas, these results suggest that the region has the potential to have nationally competitive industries in at least 12 sectors. However for this potential to be realized all of these disadvantageous features of the region will have to be corrected.

The third set of interaction terms (D^nep * D^type) is designed to check if non-state enterprises in the north-east are doing exceptionally better or worse than non-state enterprises in the rest of PRC after controlling for location and non-location factors. The coefficients on the productivity variable are always insignificant indicating no effect. However for profitability the coefficients on the interaction terms for the regional dummy with private, collective and mixed ownership dummies are negative and strongly significant, with the largest effect found for private ownership. For example, the negative coefficient of 0.0492 implies that relative to SOEs the profitability of private firms in the north-east is 4.92 percentage points lower than in the rest of the country. In terms of profitability these enterprises in the north-east seem to perform substantially worse than those in other regions, even after controlling for the negative impact of the region's general location effect. Hence private investors appear to suffer more difficulties that foreign firms, which seem better able to deal with the local business environment than the purely domestic private enterprises.

The fourth set of the interaction terms (D^nep * D^year) is designed to examine the timing of any change in performance of enterprises in the north-east relative to the rest of the country. The positive significant coefficients after 1997 for productivity and after 1998 for profitability indicate that there may have been some time impact, perhaps related to policy shifts on the speed of privatization in the later 1990's.

In regression model 4, the basic model 1 is expanded to include dummies for all 28 provinces (D^place2). The aim here is to derive a ranking of the performance gap across all provinces that is specific to location effects in each province, in another words taking away the systematic impact from factors like scale, technology, ownership, competition and sector characteristics that are captured by the control variables. The coefficients on the province dummies give the criteria for ranking. In this analysis the reference province is Shandong, so that all provinces are ranked relative to Shandong.

The results are shown in Table 7 [ PDF 87.7KB | 3 pages ]. The negative coefficients for the three north-east provinces show that their productivity and profitability is always below Shandong. For example by profitability Liaoning is 8.1 percentage points below, whilst Jilin and Heilongjiang are roughly 6.9 percentage points below. In terms of national ranking out of all provinces the three northeast provinces have three out of the bottom four places in value added per worker. By profitability the ranking out of 28 provinces is 21 for Heilongjiang, 22 for Jilin, and 27 for Liaoning. These rankings are lower than might be thought from a direct reading of data such as that in Table 1 [ PDF 87.4KB | 1 page ]and Table 2 [ PDF 55.9KB | 1 page ]and take account of both characteristics of enterprises and the structure of production in the different provinces.

We discuss the interpretation of these results further below but they show that there is a long way to go for the three provinces in improving their business environment. In other words even if they brought their situation up to the national level in terms of ownership and competition, they would still have substantially lower enterprise profitability due to their locational disadvantages.

Download this Discussion Paper [ PDF 168.5KB| 25 pages ].

Comment(s)

There are [0] comment(s) for this entry. Post a comment.