Has a private ctor emerged in China's industry?Evidence from a quarter of a million Chine firms
Sean DOUGHERTY a,⁎,Richard HERD a,⁎,Ping HE b
a OECD Economics Department,2rue AndréPascal,75775Paris Cedex 16,France
b China National Bureau of Statistics,57Yuetan Nanjie,Beijing 100826,China
Abstract
This paper asss the progress of China's transition towards a market economy by examining the structure of ownership,productivity,and profitability,as well as the concentration of production across firms,industries,and regions.It does this by analyzing a databa of firm microdata of the quarter of a million industrial companies in operation during the 1998–2003period.Results show that the private ctor now accounts for more than half of industrial output,compared with barely more than a quarter of it in 1998,and operates much more efficiently than the public ctor.Higher productivity has fed through to improved profitability,motivating greater regional specialization of production.The changes are consistent with what would be expected in a market-bad economy and suggest that reforms are making rapid progress.
©2007Elvier Inc.All rights rerved.
JEL classification:D4;F15;L11;O12;P23
“I watched a snail crawl along the edge of a straight razor …crawling and surviving ”—Francis Ford Coppola
1.Introduction
China's transition from a planned towards a market economy over the past 25years is in many ways remarkable.Highly unorthodox economic reforms were pursued in a decentralized manner,in keeping with political concerns,but often at the risk of creating new
lf-propagating China Economic Review 18(2007)309–
334
⁎Corresponding authors.Tel.:+33145248200;fax:+33144306165.
E-mail address:Sean.DOUGHERTY@oecd (S.Dougherty),Richard.HERD@oecd (R.Herd).
1043-951X/$-e front matter ©2007Elvier Inc.All rights rerved.
doi:10.1016/j.chieco.2007.02.006
310S.Dougherty et al./China Economic Review18(2007)309–334
Table1
Share of transactions conducted at market prices
197819851991199519992003 Producer goods
Market prices01346788687.3 State guided0231864 2.7 State fixed1006436161010.0 Retail sales
Market prices33469899596.1 State guided0191021 1.3 State fixed97472194 2.6 Farm commodities
Market prices64058798396.5 State guided2232047 1.6 State fixed933722179 1.9 Percent of transaction volume.
Source:China National Development and Reform Commission,Price Yearbooks and Lardy(2002). distortions.Moreover,many reforms were piecemeal and tentative at first and only accelerated gradually.The uneven path of the reforms led to well-founded concerns that the progression to a market economy could be derailed.1
少先队的呼号This paper analyzes a new t of data that covers the most recent5years of reforms through2003. The datat covers just over a quarter of a million unique firms that report their principal financial and economic results to the government on an annual basis.As a result of firm exit and entry,the number of operating firms in any given year is considerably less,amounting to between160,000 and200,000firms.The information covers all industrial companies with annual sales of over 5million yuan,and so reprents a detailed insight into the development of the Chine economy.
Private ownership was not an early feature of the reform process.Instead,local government owned e
nterpris were the early driver of reforms in the industrial ctor through the1980s and early1990s.The enterpris were marked by their ambiguous property rights which made up for a weak regulatory environment and offered investors some protection from expropriation.2Overt private ownership on a significant scale has only emerged in the past decade(OECD,2005).
In the abnce of private ownership,downstream prices were liberalized(Table1),under a dual-track system which allowed production over a t quota to be sold at market prices.While strengthening incentives,this partial deregulation of downstream prices with upstream prices t at relatively low levels left open large rents.While this approach built support for reform,its partial nature propagated distortions,in an incomplete regulatory environment with few rules governing competition behavior.
党员学习培训Concern about the and related distortions gathered pitch as appraisals of the reforms of the late1980s and early1990s revealed disturbing trends in the structure of production,prices, productivity,and trade.In a highly influential study,Alwyn Young(2000)found that the shares of different industries across provinces were converging rather than diverging(through1997),as 1Murphy,Shleifer,and Vishny(1992)illustrate the risks of getting the reform quencing wrong,such as by introducing privatization before adequately liberalizing prices,while Lau,Qian,and Roland(2001)show that the Chine reforms followed an apparently incentive-compatible path that s
ustained reform momentum even while creating(and then destroying)rent-eking opportunities along the way.And e Maddison(1998)for a long view.
2The mechanics of the unorthodox reforms are laid out by Li(1996)and Che and Qian(1998).Empirical support can be found in Dougherty and McGuckin(2002).
would be expected if comparative advantage were playing a strong role in determining the structure of production.Young also obrved that prices for industrial and retail goods were diverging rather than converging and marginal productivity was not converging across ctors,as would be expected if market forces were at work.With respect to trade,Sandra Poncet (2003,2005)found that inter-provincial barriers were having a marginally increasing negative impact on regional trade flows in the 1992to 1997period,whether looked at in aggregate or at the industry level.3
Even if one accepts the implications of Young and Poncet's studies for the period through 1997,further analysis using a variety of methods suggests that they may have reverd as time has pasd.In terms of output shares of detailed industries,Bai,Du,Tao,and Tong (2004)find that measures of regional specialization appear to have bottomed-out in the early 1990s and improved in the period through 1998.In terms of prices,Fan and Wei (2003)and Xu and V oon (2003)examine the
stationarity and co-movement of up-and downstream prices and find relatively strong evidence of convergence (or domestic integration)through 2000.4And in terms of marginal productivities,Zhang and Tan (2004)obtain evidence that the may have converged in the period through 2001.While a new input –output table is not available yet to update Poncet's analysis,direct survey evidence by the Development Rearch Center (Li,Liu,&Chen,2004)suggests that inter-provincial barriers have dropped dramatically in recent years.
We proceed first by asssing the ri of the private ctor,cond by examining its produc-tivity relative to the publicly-controlled ctor,third by decomposing the turn-around in prof-itability into subcomponents,and finally by measuring the extent of reallocation of production and conquential increa in specialization that the ownership transformation has stimulated.This empirical analysis suggests that the ri of private business has gone a long way toward turning China (starting with her industrial ctor)into a market economy.In short,the reversal of the formerly distortion-laden path appears to be decisive and is being led by widespread —and healthy (i.e.in terms of profits)—privatization of the economy,which is in turn leading to improved allocation of production and employment.
2.Data and measurement
2.1.Uniquely comprehensive microdata
The empirical work carried out in this paper utilizes the industrial firm databa of the China National Bureau of Statistics (NBS).The data cover the 1998to 2003period and include all industrial enterpris with annual sales in current yuan of 5million or higher.5Ten percent of firms are excluded in each year from the analysis since they have implausible or zero values for gross output,intermediate inputs,employment or fixed asts,bad on criteria adapted from Geng (2004).In 1998,this leaves 140,000firms with valid data,with the number rising to 180,000by 2003.As a result of exit and entry to the databa,about 80%of the firms in a given 3
Carsten Holz (2006)has recently directly contested Young's main findings and their interpretation —particularly the extent to which they imply rising trade barriers in the mid-1990s.Naughton (2003)also found less evidence of barriers between Chine provinces than Poncet in the pre-1992period.
4Byström,Olofsdotter,and Söderström (2005)also find that interregional integration has incread significantly,judged by co-movement of output shares in the decade to 2001.Viewed in a similar light,Zax and Yin (2005)find that industrial factor prices had esntially reached a plateau in terms of convergence by the 1999–2002period.5In principle the data also cover state-owned enterpris ev
en if they do not meet the threshold,but this small number of firms were excluded from the analysis.Among all industrial firms below the threshold,we estimate that the share of output by the private ctor exceeds 90%(OECD,2005).311
S.Dougherty et al./China Economic Review 18(2007)309–334
312S.Dougherty et al./China Economic Review18(2007)309–334
year have a valid obrvation in the previous year.In order to utilize the maximum number of firm obrvations,the(unbalanced)panel consisting of all firms with valid data is ud in the analysis that follows.
2.2.Financial variables adjusted to be comparable as possible
For each firm,detailed balance sheet data are available(summary shown in Table2),in addition to basic information on its ownership structure,industry,location,and employment.This high level of detail allows for veral types of adjustments to be made to value added, intermediate inputs,profit,and capital concepts to correspond more cloly with international practice.For instance,Chine statistical practice includes V AT in value added and interest in intermediate inputs;
we exclude them.In computing rates of returns on asts,the preferred profit concept is a national accounts one where bank debt and equity are treated neutrally.Thus,we add interest paid to profits prior to tax and investment income(esntially net surplus).And in measuring capital,we take the book value of net fixed asts plus inventories as an estimate.One drawback to the databa is that asts are valued at historic cost and are not revalued.However, over the period1998to2003,the increa in the price index of fixed ast investment was less than1%per year,limiting the potential bias from this source.
Since the data are firm obrvations,we are able to reclassify firms and their data items bad on firm-level ownership structure.This allows us to re-apprai previous estimates of the extent of private ownership in China.Moreover,in the final ction where we compute concentration indices,firm-level heterogeneity is taken into account in the computation of the Hoover–Glaer specialization index and Herfindahl–Hirschman(HHI)concentration ratios can be computed,on the same basis as the U.S.Census Bureau.
3.Ownership
3.1.Definition of private control
According to contemporary theories of the firm,ownership should be defined in terms of which shareholder controls the“residual rights”of the firm,in the n of who dictates unforeen contingencies(Hart,1995).This is the definition that we ek to apply.Since detailed data on the shareholding structure of each firm is available,the type of controlling shareholder can be identified.This approach contrasts with the official firm registration categories that are commonly ud in analysis of China's business ctor.
Rather than using the official firm registration status that is commonly shown in Chine statistical publications to look at ownership,we parate firms by type of controlling shareholder. By examining shareholding structure,firms can be parated by whether it is the state(directly or indirectly through legal persons),a collective(local government),or a private entity(individuals, domestic legal persons,or foreign companies)that controls the firm.This classification allows us to look at the type of actual owner,since the official registered enterpri structure often does not reflect the type of owner that controls the firm becau companies rarely change their registration status,even when their controlling shareholder changes.The u of shareholder information is especially important for limited liability and shareholding companies(dominated by legal person shareholders),who ownership is heavily mixed between state and private control.
杀鸡焉用宰牛刀In the NBS microdata,firms directly report whether they are state held:that is,they are controlled directly or indirectly by the state.In order to distinguish between direct and indirect
Table 2
Profile of industrial microdata (all firms meeting criteria)a
1998
19992000200120022003Basic data
Units Number of firms
144,112140,559141,956150,948160,969180,146Millions of persons Employment
59.255.553.051.752.655.5Billion yuan Value added (including V AT)
1842.82046.92413.42693.13154.34069.2Value added tax payable
270.4298.6352.0384.8431.6535.4Sales tax and extra charges
121.2128.3140.4151.9172.3202.4Gross output comparable (1990)prices
4946.15242.96170.06861.97978.510,345.4Gross output current prices
6518.56991.18231.49170.210,701.213,879.1Intermediate input
5556.76260.07300.78365.49965.612,591.8Exports
1049.41124.51413.81587.51964.62647.0Income statement
Billion yuan Sales Revenue
6168.06713.58090.49008.710,577.513,966.8Cost of sales
5072.55499.16589.27405.68692.711,557.6Sales charge
219.4248.1288.7338.3394.2484.4Sales tax and extra charge
121.2128.3140.4151.9172.3202.4Sales profit
754.9838.11072.21112.91318.21722.4Other business profits
41.336.345.047.952.369.8Administrative fees and charges时间太快
466.6487.0541.9572.6640.9780.1Memo:R&D
–––26.234.943.8Financial charges
232.5210.3188.9181.6192.7206.9Memo:interest outlay
214.4194.5177.8168.2172.8181.7Operating profit
97.1177.1386.4406.6536.9805.1Subsidies
翅尖怎么做好吃27.627.532.535.138.346.5Investment income and adjustments
−16.7−19.2−8.4−21.2 4.530.1Profit b
141.4223.8427.3462.9570.7821.5Tax on profits
51.160.674.987.0108.6141.3Dividends
56.263.087.696.2114.3–Retained earnings
34.0100.1264.9279.7347.8–Balance sheet
Billion yuan Asts
7367.97843.48291.38902.09569.910,947.8Net fixed asts
4117.44516.44909.85231.75657.06406.4Inventories
1470.91470.31557.01642.81743.22040.4Deferred and intangible asts
1779.61856.71824.62027.42169.72501.0Liabilities
7367.97843.48291.38902.09569.910,947.8Long term liabilities
1951.32005.52065.82053.42119.12319.3Net short term liabilities
1612.21547.41472.91506.31585.11787.9Equity
3768.94241.54694.35280.75816.36755.8Performance indicators
Profit plus interest c as a percent of fixed asts plus inventories Rate of return on physical asts
6.1 6.79.28.910.112.2Profits as a percent of equity Rate of return on equity
3.8 5.39.18.89.812.2Billion yuan Profit plus interest c
339.1399.1596.7609.8748.01033.3Net fixed asts plus inventories
5588.35986.76466.86874.57400.28446.8Percent of net fixed asts Depreciation rate –8.18.18.58.79.3Source:China National Bureau of Statistics with joint NBS-OECD analysis.
a Missing obrvations and tho with inconsistent data are not included.
b Referred to as “Total Profits ”in most Chine publications.
沂水春风c Excludes investment income.
313S.Dougherty et al./China Economic Review 18(2007)309–334
理财管理314S.Dougherty et al./China Economic Review18(2007)309–334
control,firms where the state owns more than50%of the share capital6are classified as directly state controlled,with the remainder of state-held firms treated as indirectly controlled(by legal persons or other shareholders).Amongst the non-state held firms,collective controlled firms are identified if they
report collective capital share greater than50%.The remainder of the non-state firms are subdivided among various types of private ownership depending on whether they are controlled(share capital greater than50%)by a private company(a non-state legal person), individuals,non-mainland agents,or other shareholders.This classification by controlling share-holder—which is exhaustive—allows us to look at the type of actual owner,since the official registration status often does not reflect the de facto owner.The u of controlling shareholder also overcomes the difficulties in interpreting the bewildering array of different ownership registration categories,many of which are not meaningfully distinct.See OECD(2000)and ADB(2003)for a detailed list of the legal basis for each enterpri type.
3.2.Results
The result of this exerci is to show a rapid shift toward private ownership in China amongst firms with more than5million yuan in annual sales(Table3).Classification by controlling shareholder shows that the private ctor has grown from27.9%of industrial value added in1998 to52.3%in2003.Individually-controlled firms'share has grown most rapidly,reprenting almost half of this increa,with the remainder of the gain split equally between private companies(controlled by non-state legal persons)and non-mainland shareholders.Part of this increa may be attributed to th
e progressively larger share of all firms that have had to report to statistical authorities over the years,since most of the firms that have crosd over the reporting firm size threshold appear to be private.7
At the same time as the private ctor has grown,the state and collective controlled ctors have fallen.The share of value added directly controlled by the state fell from38.9%to22.9%over the 5years1998–2003.While veral percentage points of this drop may reprent a shift toward indirect state control,the remainder occurred through the closure,restructuring,and privatization of enterpris.The collective controlled share has also fallen rapidly,as many of the firms have also exited or changed ownership.
Defining ownership bad on control can differ significantly from the registered ownership categories that are usually relied upon.For instance,more than half of limited liability companies' value added was controlled by private shareholders in2003,as was a quarter of shareholding companies and joint ownership enterpris'value added.Many such corporations were formerly state controlled and since have been privatized.In many studies,such companies are assumed to either lie completely in the public or the private ctors(e ADB,2003;IFC,2000;Song,Tenev, Yao,&Garnaut,2005).
偶然拼音
Even within registered ownership categories that should be more homogenous,there is a range of controlling interests.For example,although all structures classified as state-owned enterpris and solely state-funded corporations are controlled by the state in some way,over20%were indirectly controlled by other state controlled companies(through legal person shares).And while 6There are drawbacks to such a classification.Share ownership is only one of the criteria that the propod revision of the System of National Accounts(SNA)suggests as relevant.In particular,the SNA suggests looking at groups that control the board of directors.It may be possible to exert control through the board with less than50%of the share capital.
7Nevertheless,a generalization of the results to the whole commercial business ctor and to aggregate GDP confirms a rapid shift toward private ownership,with it overtaking public ownership during the past5years(OECD,2005).
