中國(guó)的城鄉(xiāng)收入差距和不平等外文翻譯（節(jié)選）

1、<p>　　3950英文單詞，2.5萬(wàn)英文字符，中文6750字</p><p>　　文獻(xiàn)出處：Sicular T, Yue X, Gustafsson B, et al. THE URBAN–RURAL INCOME GAP AND INEQUALITY IN CHINA[J]. Review of Income & Wealth, 2006, 53(1):93-126.</p>

2、<p>　　THE URBAN–RURAL INCOME GAP AND INEQUALITY IN CHINA</p><p>　　Terry Sicular,Ximing Yue,Björn Gustafsson,and Shi Li</p><p><b>　　Abstract</b></p><p>　　Using

3、new household survey data for 1995 and 2002, we investigate the size of China’s urban-rural income gap, the gap’s contribution to overall inequality in China, and the factors underlying the gap. Our analysis improves on

4、past estimates by using a fuller measure of income, adjusting for spatial price differences and including migrants. Our methods include inequality decomposition by population subgroup and the Oaxaca- Blinder decompositio

5、n. Several key findings emerge. First, the adjustments</p><p>　　Keywords: China, urban-rural inequality, income </p><p>　　1. Introduction </p><p>　　Studies of China’s inequality alm

6、ost universally report that the gap between urban and rural household incomes in China is large, has increased over time, and contributes substantially to overall inequality. According to most estimates mean per capita i

7、ncome in urban China is more than triple that in rural areas, giving China one of the highest urban-rural income ratios in the world. The size of this gap has been discussed in the Chinese official media, is noted in gov

8、ernment and communist party</p><p>　　China’s urban-rural income gap is often attributed to policies that have inhibited labour mobility, most importantly the household registration or hukou system. The house

9、hold registration system was established during the Maoist period to control population movement. It has continued to the present and is reinforced by a range of complementary policies such as taxation of urban employers

10、 that hire migrants, prohibition of urban employment of migrants in some trades, and the denial of urban publi</p><p>　　Whether or not concerns about the urban-rural income gap are justified depends, among o

11、ther things, on the true magnitude of the gap as well as on the factors that underlie the gap. To date a range of studies have examined China’s urban-rural income gap (e.g., Knight and Song 1999; Yang and Zhou 1999; Zhao

12、 and Tong 2000; Shi 2004; Sicular et al. 2004; Benjamin et al. 2005). For several reasons, most linked to data constraints, past estimates of the income gap are likely biased. </p><p>　　First, most income da

13、ta for China do not include certain components. One missing component is housing-related income, specifically, the imputed rental value of owner- occupied housing and imputed subsidies on publicly-owned rental housing. H

14、ousing- related income is likely to differ systematically between urban and rural areas, and it is of recent interest because in the late 1990s China privatized urban housing (Khan and Riskin 2007). Also missing from inc

15、ome is the value of household consumpti</p><p>　　With these considerations in mind, here we recalculate the size of China’s urban-rural income gap, estimate the contribution of the revised gap to overall ine

16、quality, and analyze the factors underlying the gap. For our analysis we use household and individual data from household income surveys for 1995 and 2002. These surveys were conducted under the auspices of the Chinese A

17、cademy of Social Sciences (CASS). They are large, nationally representative, and contain detailed information on household</p><p>　　The first step of our analysis is to recalculate the size of the urban-rura

18、l gap and its contribution to national inequality. We do so for China as a whole and for its three major regions—the east, center and west. In this recalculation we make three modifications that bring our measurement of

19、the gap closer to international best practice and allow more comparability with studies for other countries. </p><p>　　First, we use a fuller measure of income that includes housing-related components of inc

20、ome. Unfortunately, we cannot measure the implicit subsidies associated with household consumption of public services. From a theoretical standpoint this should be included to fully capture urban-rural differences, but t

21、he necessary information is unavailable. More generally, data on household consumption of public services is unavailable for most countries and rarely included in international calculations of</p><p>　　The s

22、econd step of our analysis is an investigation of the factors underlying the urban- rural gap. Here we use the Oaxaca-Blinder decomposition. The Oaxaca-Blinder method cannot identify how particular policies such as the h

23、ousehold registration system contribute to the gap, but it gives information on the extent to which the gap reflects differences between urban and rural areas in household characteristics as opposed to simple location of

24、 residence. This method also gives a measure, albeit fr</p><p>　　Several key findings emerge from our analysis. We find that after recalculation, the urban-rural income gap is substantially reduced. While in

25、cluding housing-related income components increases the income gap somewhat, adjusting for spatial price differences dramatically reduces it. Including migrants narrows the gap further. With these revisions, China still

26、has a relatively large urban-rural income ratio, but that ratio is within the range of most other countries. It follows that these adjustm</p><p>　　The Oaxaca-Blinder decomposition reveals that household and

27、 individual characteristics such as education, age, and household demographics, indeed contribute to the urban- rural income gap. Differences in the endowments of such characteristics, holding the returns to these charac

28、teristics constant, contributed about half of the income gap. Location of residence, including differences between urban and rural areas in the returns to household and individual characteristics, contributed the other h

29、a</p><p>　　2. Definitions and data </p><p>　　The data used for the analysis in this paper come from two rounds of the CASS Household Income Survey conducted in 1996 and 2003 for the reference pe

30、riods of 1995 and 2002. These surveys were carried out under the direction of a team of researchers consisting of scholars at the Institute of Economics, Chinese Academy of Social Sciences, and researchers from other cou

31、ntries. The data were collected by the NBS using survey instruments designed by the project research team. A detailed description o</p><p>　　Regional coverage changed somewhat between the two years of the su

32、rvey. To ensure comparability between the results for the two years, we use a subsample having the property that each location (province*rural, province*urban) was present in the survey for both years under investigation

33、. The rural sample covers Anhui, Beijing, Gansu, Guangdong, Guizhou, Hebei, Henan, Hubei, Hunan, Liaoning, Jiangsu, Jiangxi, Jilin, Shaanxi, Shandong, Shanxi, Sichuan, Yunnan, and Zhejiang. The urban sample covers </p

34、><p>　　Since urban residents were over-sampled in 1995 and under-sampled in 2002, we weight the urban and rural subsamples so that their population shares equal those in the total population according to offici

35、al NBS census-based population data. With this adjustment, the sample distribution between rural and urban areas is consistent with the official population distribution between urban and rural areas for all of China. All

36、 analyses using the combined urban and rural samples use this population weig</p><p>　　A limitation of most household survey data for China is that rural-to-urban migrants who do not have an urban residence

37、permit are excluded. For 2002 the CASS survey includes a special sample of migrants, making it possible to produce more complete estimates for that year. In Section 5 below we describe the migrant sample and explore how

38、including migrants influences the size of the rural-urban gap and its contribution to inequality. Section 5 also contains a more general discussion of China’s u</p><p>　　The target variable for this study is

39、 household per capita disposable income.5 This includes cash income, retained in-kind income (important in rural China, particularly at the beginning of the period studied), and other income in kind (relevant in urban Ch

40、ina in the past, although declining in importance in recent years). Net taxes and fees are subtracted. </p><p>　　Most economists believe that income should include housing-related components. The NBS does no

41、t include these components in disposable income, nor do most other household income data for China. Our estimates of average household income in China use the NBS definition but add in housing subsidies and imputed rent.

42、 Income levels here are therefore higher than those obtained using the NBS definition. Depending on the distributional profile of housing subsidies and imputed rent, our definition of in</p><p>　　During the

43、period under investigation, China carried out housing reform in urban China. In the past most urban households had lived in public housing and paid low rent, implying that they received rental housing subsidies. These su

44、bsidies largely benefited better-off households (Khan et al. 1993). In the late 1990s the government privatized urban public housing. By 2002 most urban residents owned their homes and no longer received rental housing s

45、ubsidies. Rather, they now received the imputed r</p><p>　　Our analysis treats the household as the income-receiving unit. Disposable income of each household is then divided by the number of household membe

46、rs. Following what is now common practice in analysis of income distributions, we assign this household average to each member of the household. Individuals are thus the unit of analysis, and we abstract from intra-house

47、hold allocation issues. </p><p>　　Since price levels have changed over time, and differentially among provinces and between rural and urban areas, we use official provincial consumer price indices to express

48、 2002 incomes in 1995 prices. Note that separate indices are available for rural versus urban areas in each province. We use these separate indices so that deflation factors can differ between urban and rural areas withi

49、n provinces as well as among provinces. Prices differ not only across time, but also spatially at any point</p><p>　　The Brandt-Holz spatial price deflators have some limitations. One is that their estimates

50、 of housing costs are based on the costs of housing construction materials, and the difference in the costs of construction materials between urban and rural areas is typically smaller than the difference in costs of hou

51、sing services. For this reason, the Brandt-Holz estimates may understate the price differential between urban and rural areas. Also, they only have raw price data for 1990, and they use a ba</p><p>　　Conclus

52、ion </p><p>　　In this paper we have explored China’s urban-rural income gap. Several key findings emerge. First, China’s urban-rural income gap is large by international standards, even after various adjustm

53、ents such as fuller measurement of income, spatial price deflation and including migrants in the urban sample. Still, these adjustments, especially spatial price deflation, reduce the size of the gap substantially. With

54、respect to trends over time, we find the adjusted relative gap widened little between 1</p><p>　　Second, the contribution of the urban-rural income gap to overall inequality has been relatively large and has

55、 increased somewhat, although again its level is reduced by the adjustments. If we use 2002 unadjusted incomes and include migrants, between-group inequality contributes more than 40 per cent of overall inequality. If we

56、 further correct for spatial price differences, the contribution declines to 26 per cent. With or without adjustments, this contribution is large relative to that in othe</p><p>　　Third, regional differences

57、 in China’s urban-rural gap are large. The urban-rural income gap is much larger in western China than in the eastern or central regions, as is its contribution to inequality. Indeed, the urban-rural gap’s contribution t

58、o overall inequality in the east and center is fairly small. These regional differences suggest that efforts to bridge the urban-rural divide should target the west. Further research is required to identify what sorts of

59、 targeted interventions would be </p><p>　　Our analysis highlights several measurement issues. One is spatial differences in prices and the cost of living, which has a substantial impact on the measured size

60、 of the gap and its contribution to inequality. Our findings here parallel those in Brandt and Holz (2004). As the study of income inequality is ultimately interested in real differences in incomes, spatially adjusted es

61、timates of the urban-rural gap and its contribution are most meaningful. </p><p>　　A second measurement issue is the delineation of urban versus rural populations. Here various problems arise, but probably m

62、ost important for China is the treatment of migrants. Including migrants in the urban sample reduces the size of the urban-rural income gap and that gap’s contribution to inequality, but only modestly. Including migrants

63、 has little impact on the overall level of inequality, because lower between- group inequality is offset by higher within-group inequality. Migration increa</p><p>　　Further research and better data are need

64、ed to fully explore the impact on inequality measurement of including migrants, but these results provide some indication of the magnitude and direction. The impact is, however, noticeably smaller than that of correcting

65、 for spatial price differences. Efforts to improve information on geographic price differences, then, are equally important. </p><p>　　What explains the urban-rural gap? Differences in endowments of househol

66、d characteristics contribute roughly half the gap in PPP ln income. Most important here is education. Differences in education levels contribute 25-30 per cent of the gap. These estimates imply that, all else equal, if r

67、ural education levels were increased to be on a par with urban levels, the urban-rural income gap would decline by 25-30 per cent. </p><p>　　Location of residence contributes the other half of the PPP income

68、 gap. Here location’s contribution is defined as the sum contribution to the gap of differences between urban and rural areas in the constant terms, coefficients on provincial dummy variables, and coefficients on househo

69、ld and individual characteristics. Spatial price deflation makes a difference here, reducing location’s contribution by more than 10 percentage points. Over time the contribution of location declines somewhat, whic</p

70、><p>　　Our analysis points to the need for further research in several areas. One is education. Studies on education in China generally report large differences in the levels of education not only between urban

71、 and rural areas but also among provinces (Hannum etal. 2005) Such spatial differences in education likely reflect multiple factors, including differences in incomes, in public expenditures on education, and in patterns

72、of migration. Evidence provided here and elsewhere suggests that the private </p><p>　　A second topic for further research is spatial location. Why, after controlling for observed characteristics, does locat

73、ion of residence remain so important in explaining income differences? The hukou or household registration system and related policies that continue to hinder rural-to-urban movement are obvious culprits. Yet the persist

74、ence of urban-rural gaps in other countries suggests that even without such artificial restrictions, migration is unlikely to eliminate the urban-rural income g</p><p>　　China’s urban-rural income gap has sh

75、own little sign of declining despite substantial easing of the restrictions on migration and the growing number of migrants. A variety of factors could contribute to the persistence of spatial differences. One factor is

76、non- labour income, which accounts for nearly half the income gap. Migration is not likely to reduce gaps in some forms of non-labour income such as housing-related income and pensions. Also, migration may not be able to

77、 eliminate the gap beca</p><p>　　中國(guó)的城鄉(xiāng)收入差距和不平等</p><p>　　摘要：我們利用1995年到2012年的住戶數(shù)據(jù)，調(diào)查了中國(guó)城鄉(xiāng)收入差距的大小、差距對(duì)中國(guó)整體不平等的貢獻(xiàn)以及差距的因素。我們的分析與過(guò)去研究相比，采用了更為全面的收入數(shù)據(jù)來(lái)計(jì)算，并調(diào)整了地區(qū)差異以及遷移因素也有考慮。我們的方法包括人口子群體的不等式分解和Oaxac-Blind

78、er分解。本研究有幾個(gè)重要的發(fā)現(xiàn)：一是調(diào)整后大幅度降低了中國(guó)城鄉(xiāng)收入差距及其對(duì)不平等的貢獻(xiàn)，盡管如此，差距仍然很大，并且隨著時(shí)間的推移有所增加。二是在控制家庭個(gè)性化因素之后，居住區(qū)位仍然是影響城鄉(xiāng)收入差距的最重要因素。我們發(fā)現(xiàn)唯一有助于縮小差距的家庭個(gè)性因素是教育，其他家庭因素的差異是相對(duì)不重要的。</p><p>　　關(guān)鍵詞：中國(guó)，城鄉(xiāng)不平等，收入</p><p><b>　　

79、引言</b></p><p>　　中國(guó)不平等現(xiàn)象的研究幾乎普遍地表明，中國(guó)城鄉(xiāng)家庭收入差距很大，隨著時(shí)間的推移不斷增加，對(duì)全面的不平等狀況起了重大作用。據(jù)大多數(shù)人估計(jì)，中國(guó)城鄉(xiāng)人均收入是農(nóng)村人均收入的三倍以上，使中國(guó)成為世界城鄉(xiāng)收入比例最高的國(guó)家之一。這個(gè)差距的大小已經(jīng)在中國(guó)的官方媒體中得到討論，在政府和共產(chǎn)黨的報(bào)告中得到了注意，也是近期主要政策舉措的動(dòng)力，如“建設(shè)社會(huì)主義新農(nóng)村”運(yùn)動(dòng)，旨在縮小差距通

80、過(guò)增加農(nóng)村公共支出。</p><p>　　中國(guó)的城鄉(xiāng)收入差距往往歸因于抑制勞動(dòng)力流動(dòng)的政策，最重要的是戶籍制或戶籍制。毛澤東時(shí)期建立了戶籍制度，以控制人口流動(dòng)。它一直延續(xù)到現(xiàn)在，并通過(guò)一系列補(bǔ)充政策來(lái)加強(qiáng)，例如雇用遷移的城鎮(zhèn)雇主征稅，禁止一些行業(yè)的城市就業(yè)遷移，以及拒絕城市公共服務(wù)，如對(duì)未登記的家庭的教育。近年來(lái)，政府對(duì)戶籍體制進(jìn)行了改革，使流動(dòng)性更大，但存在巨大障礙（Wang 2004）。這些障礙被認(rèn)為是保護(hù)城

81、市居民的福利，這是一個(gè)政治敏感的群體，但同時(shí)又引起了不同類(lèi)型的政治問(wèn)題。</p><p>　　對(duì)城鄉(xiāng)收入差距的擔(dān)憂是否合理取決于差距的真實(shí)程度以及構(gòu)成差距的因素。迄今為止，一系列研究已經(jīng)檢驗(yàn)了中國(guó)城鄉(xiāng)收入差距（如Knight和Song 1999; Yang and Zhou 1999; Zhao and Tong 2000; Shi 2004; Sicular等2004; Benjamin等2005）。由于幾個(gè)原

82、因，大多數(shù)與數(shù)據(jù)約束有關(guān)，過(guò)去對(duì)收入差距的估計(jì)可能有偏差。</p><p>　　首先，中國(guó)大多數(shù)收入數(shù)據(jù)不包括某些組成部分。一個(gè)缺失的部分是住房相關(guān)收入，特別是自住房屋的估算租金價(jià)值和對(duì)公有租賃房屋的估算補(bǔ)貼。住房相關(guān)收入在城鄉(xiāng)之間可能有系統(tǒng)的差異，因?yàn)樵?0世紀(jì)90年代后期，中國(guó)城鎮(zhèn)住房私有化（Khan and Riskin，2007）。收入中也缺少家庭消費(fèi)在教育，保健和地方基礎(chǔ)設(shè)施等領(lǐng)域的價(jià)值。公共服務(wù)的消費(fèi)

83、在城市中可能會(huì)比農(nóng)村更高，因此排除這一情況會(huì)導(dǎo)致中國(guó)城鄉(xiāng)差距的低估。第二，大多數(shù)研究不能控制生活費(fèi)用的空間差異。這是可以理解的，因?yàn)榭臻g價(jià)格差異的系統(tǒng)信息很少。然而，如果城市生活成本明顯高于農(nóng)村，那么收入差距可能比文獻(xiàn)報(bào)道的要差。第三，大多數(shù)中國(guó)城鄉(xiāng)收入差距的估計(jì)數(shù)據(jù)都是根據(jù)不存在城市居民的未登記遷移的數(shù)據(jù)。由于農(nóng)村到城市的遷移一般被認(rèn)為是縮小城鄉(xiāng)收入差距的重要機(jī)制，不包括農(nóng)村到城市的遷移是有問(wèn)題的。不包括中國(guó)的平均收入低于農(nóng)村居民的收

84、入，導(dǎo)致城鄉(xiāng)收入差距過(guò)高。包括遷移在中國(guó)的研究越來(lái)越重要，因?yàn)檫w移的限制已經(jīng)松動(dòng)，遷移相應(yīng)增加。</p><p>　　考慮到這些因素，在這里我們重新計(jì)算了中國(guó)城鄉(xiāng)收入差距的大小，估計(jì)了修正后的差距對(duì)總體不平等的貢獻(xiàn)，并分析了差距的因素。對(duì)于我們的分析，我們使用1995年和2002年家庭收入調(diào)查的家庭和個(gè)人數(shù)據(jù)。這些調(diào)查是在中國(guó)社會(huì)科學(xué)院（CASS）的主持下進(jìn)行的。它們規(guī)模大，具有國(guó)家代表性，并載有關(guān)于家庭收入和其

85、他相關(guān)家庭和個(gè)人因素的詳細(xì)信息。CASS數(shù)據(jù)具有一定的優(yōu)勢(shì)。它們是相對(duì)較新的，因此提供比通?？捎玫母嗟淖钚滦畔?。其他具有廣泛區(qū)域覆蓋率的數(shù)據(jù)，特別是國(guó)家統(tǒng)計(jì)局（NBS）的家庭調(diào)查數(shù)據(jù)，通常只能以研究人員的形式列出或匯總。提供諸如中國(guó)健康與營(yíng)養(yǎng)調(diào)查（CHNS）等家庭層面調(diào)查數(shù)據(jù)的替代數(shù)據(jù)集比CASS調(diào)查的區(qū)域覆蓋范圍更窄。最后，CASS數(shù)據(jù)集是不尋常的，因?yàn)樗》肯嚓P(guān)收入組成部分和農(nóng)村到城市遷移的信息。</p><

86、;p>　　我們分析的第一步是重新計(jì)算城鄉(xiāng)差距的大小及其對(duì)國(guó)家不平等的貢獻(xiàn)。我們對(duì)中國(guó)整體和三大地區(qū)：東，中，西都是這樣做的。在這次重新計(jì)算中，我們進(jìn)行了三項(xiàng)修改，使我們對(duì)差距的衡量更接近國(guó)際最佳做法，并與其他國(guó)家的研究更具可比性。</p><p>　　首先，我們采用更全面的收入來(lái)衡量住房相關(guān)的收入組成部分。不幸的是，我們無(wú)法衡量與家庭消費(fèi)公共服務(wù)相關(guān)的隱性補(bǔ)貼。從理論的角度來(lái)看，這應(yīng)該被納入到充分發(fā)揮城鄉(xiāng)差

87、異的作用，但不能得到必要的信息。更廣泛地說(shuō)，大多數(shù)國(guó)家的公共服務(wù)消費(fèi)數(shù)據(jù)不可用，很少包括在家庭收入的國(guó)際計(jì)算中。二是調(diào)整生活費(fèi)用的空間差異。生活成本在區(qū)域間和城鄉(xiāng)之間有系統(tǒng)的差異，所以理想的不平等研究應(yīng)該是利用空間價(jià)格平價(jià)指數(shù)調(diào)整的收入。由于缺乏區(qū)域價(jià)格水平的數(shù)據(jù)，空間價(jià)格通縮在中國(guó)研究中很少見(jiàn)。一個(gè)例外是Ravallion和Chen（2004），它使用城鄉(xiāng)貧困線的估計(jì)來(lái)調(diào)整生活成本差異，然后重新計(jì)算國(guó)家不平等。最近，Benjamin等

88、 （2005）利用Brandt和Holz（2004）研究的新空間價(jià)格指數(shù)來(lái)重新計(jì)算國(guó)家不平等程度。然而，這些研究都沒(méi)有使用空間價(jià)格通縮來(lái)詳細(xì)分析城鄉(xiāng)收入差距。在這里，我們使用布蘭特和霍爾茨（同上）空間價(jià)格指數(shù)來(lái)縮小收入，然后重新計(jì)算城鄉(xiāng)收入差距及其對(duì)不平等的作用。在相關(guān)的情況下，我們將結(jié)果與Ravallion和Chen（同上）和Benjamin等人進(jìn)行了比較。（同上）。第三，我們包括農(nóng)村到城市的遷移</p><p&g

89、t;　　我們分析的第二步是調(diào)查城鄉(xiāng)差距的因素。這里我們使用瓦哈卡布林德分解。瓦哈卡盲人法不能確定戶籍制度等特定政策如何有助于彌補(bǔ)差距，而是提供差距反映城鄉(xiāng)差異在家庭因素方面的信息，而不是簡(jiǎn)單的居住地點(diǎn)。這種方法也提供了一個(gè)措施，盡管從部分均衡的角度來(lái)看，如果農(nóng)村和城市群體具有相似的因素，差距就會(huì)有多大。這樣的信息從政策的角度來(lái)看是有用的。例如，如果我們發(fā)現(xiàn)農(nóng)村和城市之間的教育因素差異大大地影響了差距，那么政策制定者可能希望將注意力集中在

90、兩個(gè)部門(mén)的教育水平的決定因素和后果上。</p><p>　　我們的分析發(fā)現(xiàn)了幾個(gè)關(guān)鍵的結(jié)果。我們發(fā)現(xiàn)，經(jīng)過(guò)重新計(jì)算，城鄉(xiāng)收入差距大大減少。雖然包括住房相關(guān)收入組成部分在一定程度上增加了收入差距，但調(diào)整空間價(jià)格差異卻大大減少。包括遷移在內(nèi)的差距進(jìn)一步縮小。經(jīng)過(guò)這些修訂，中國(guó)的城鄉(xiāng)收入比例仍然相對(duì)較大，但這個(gè)比例在其他大多數(shù)國(guó)家的范圍之內(nèi)。因此，這些調(diào)整也減少了中國(guó)城鄉(xiāng)差距對(duì)全面不平等的貢獻(xiàn)。在重新計(jì)算收入和包括遷移

91、之后，我們發(fā)現(xiàn)，在2002年，城鄉(xiāng)差距貢獻(xiàn)了總體不平等的四分之一，而在大多數(shù)研究中估計(jì)為50％以上。</p><p>　　瓦哈卡盲人分解顯示，家庭和個(gè)人因素，如教育，年齡和家庭人口，確實(shí)有助于城鄉(xiāng)收入差距。這些因素的稟賦差異使這些因素的回報(bào)不變，貢獻(xiàn)了大約一半的收入差距。居住地位，包括城鄉(xiāng)差距，家庭回報(bào)和個(gè)人因素，貢獻(xiàn)了另一半。有趣的是，1995年至2002年期間，地點(diǎn)的貢獻(xiàn)率有所下降，盡管僅有一點(diǎn)。這與這段時(shí)間

92、內(nèi)空間流動(dòng)性的增加是一致的。分解表明，教育是城鄉(xiāng)收入差距最重要的非定位因素。2002年，城鄉(xiāng)差距在教育水平上貢獻(xiàn)了四分之一的收入差距。家庭規(guī)模、組成，土地所有權(quán)和黨員身份等其他家庭因素的稟賦和回報(bào)的差異平衡起來(lái)不太重要。也就是說(shuō)，從長(zhǎng)遠(yuǎn)來(lái)看，教育水平是內(nèi)生性的，目前對(duì)人力資本的投資可能受到家庭規(guī)模和組成等其他家庭因素的影響。</p><p><b>　　2.定義和數(shù)據(jù)</b></p&g

93、t;<p>　　本文用于分析的數(shù)據(jù)來(lái)自于1996年和2003年對(duì)1995年和2002年參考期進(jìn)行的兩輪中國(guó)社會(huì)科學(xué)院家庭收入調(diào)查。這些調(diào)查是在中國(guó)社會(huì)科學(xué)院經(jīng)濟(jì)研究所的學(xué)者和其他國(guó)家的研究人員組成的研究小組的指導(dǎo)下進(jìn)行的。數(shù)據(jù)由國(guó)家統(tǒng)計(jì)局采用項(xiàng)目研究小組設(shè)計(jì)的調(diào)查工具收集。數(shù)據(jù)的詳細(xì)描述可參見(jiàn)Li等（2007年）。這里我們指出一些數(shù)據(jù)集的主要因素，并討論與我們的分析最相關(guān)的方面。</p><p>　

94、　調(diào)查兩年之間，區(qū)域覆蓋面有所變化。為了確保兩年的結(jié)果之間的可比性，我們使用一個(gè)子樣本，每個(gè)位置都在調(diào)查中存在于調(diào)查中。農(nóng)村樣本涵蓋安徽，北京，甘肅，廣東，貴州，河北，河南，湖北，湖南，遼寧，江蘇，江西，吉林，陜西，山東，山西，四川，云南，浙江。城市樣本覆蓋安徽，北京，甘肅，廣東，河南，湖北，江蘇，遼寧，山西，四川，云南。</p><p>　　由于1995年城市居民的抽樣調(diào)查和2002年抽樣調(diào)查，我們對(duì)城市和農(nóng)村

95、子樣本加權(quán)，使他們的人口在根據(jù)官方的國(guó)家統(tǒng)計(jì)局的人口數(shù)據(jù)，在總?cè)丝谥衅椒?。通過(guò)這種調(diào)整，農(nóng)村和城市之間的樣本分布與中國(guó)全國(guó)城鄉(xiāng)官方人口分布一致。所有使用城鄉(xiāng)樣本綜合的分析使用這個(gè)比重調(diào)整。</p><p>　　中國(guó)大多數(shù)家庭調(diào)查數(shù)據(jù)的局限性是，沒(méi)有城市居留證的農(nóng)村到城市的遷移被排除在外。2002年，中國(guó)社會(huì)科學(xué)院的調(diào)查包括一個(gè)特殊的遷移樣本，使得有可能在當(dāng)年進(jìn)行更完整的估計(jì)。在下面的第5節(jié)中，我們描述了遷移樣本，

96、并探索了遷移如何影響到城鄉(xiāng)差距的大小及其對(duì)不平等的貢獻(xiàn)。第5節(jié)還對(duì)中國(guó)城市人口統(tǒng)計(jì)進(jìn)行了更為一般的討論。</p><p>　　本研究的目標(biāo)變量是家庭人均可支配收入，其中包括現(xiàn)金收入，留存實(shí)物收入（特別是在中國(guó)農(nóng)村，特別是研究開(kāi)始時(shí)），以及其他實(shí)物收入（在中國(guó)城市相關(guān)過(guò)去雖然近年來(lái)的重要性不斷下降）。大多數(shù)經(jīng)濟(jì)學(xué)家認(rèn)為，收入應(yīng)包括與房屋有關(guān)的組成部分。國(guó)家統(tǒng)計(jì)局不包括這些可支配收入的組成部分，也不包括中國(guó)大部分家庭