Urbanization, Structural Transformation and Rural-Urban Disparities in India and China

30 Pages·2015·0.52 MB·English

Urbanization, Structural Transformation and
Rural-Urban Disparities in India and China
Viktoria Hnatkovska and Amartya Lahiri
(cid:3) y
April 2015
Very Preliminary
Abstract
Over the past three decades India and China have experienced rapid economic growth along
witharapidstructuraltransformationoftheeconomyawayfromagriculturetowardsmanufactur-
ingandservices. Underneaththeoverallsimilarityinaggregatetakeo⁄showeverisonesigni(cid:133)cant
di⁄erence: rural and urban disparities have evolved very di⁄erently in the two countries: rural-
urban wage gaps have been declining in India while they have widened in China during this
period. We formalize a two-sector model of structural transformation where rural-urban wage
gaps depend on two factors: (i) di⁄erential sectoral income elasticities of demand along with
productivity growth; and (ii) labor supply growth in urban areas due to urbanization. While the
modelcanquantitativelyaccountforthewageconvergenceinIndia,theresultsforChinaindicate
the possibility of ine¢ cient urbanization there.
JEL Classi(cid:133)cation: J6, R2
Keywords: Rural urban disparity, urbanization, structural transformation
(cid:3)DepartmentofEconomics,UniversityofBritishColumbia,997-1873EastMall,Vancouver,BCV6T1Z1,Canada
and Wharton School, University of Pennsylvania. E-mail address: [email protected].
yDepartmentofEconomics,UniversityofBritishColumbia,997-1873EastMall,Vancouver,BCV6T1Z1,Canada.
E-mail address: [email protected].
1
1 Introduction
The process of economic development typically involves large scale structural transformation of
economies. As documented by Kuznets (1966), structural transformations typically involve a con-
traction in the agricultural sector accompanied by an expansion of the non-agricultural sectors (cid:150)
manufacturingandservices. Inasmuchasthecontractingagriculturalsectorisprimarilyruralwhile
the expanding sectors mostly urban, the structural transformation process has potentially important
implications for the evolution of economic inequality within such developing economies. The process
clearly induces large reallocation of workers across sectors as well as requires, possibly, re-training
of workers to enable them to make the switch. Not surprisingly, in a recent cross-country study on
a sample of 65 countries, Young (2012) (cid:133)nds that around 40 percent of the average inequality in
consumption is due to urban-rural gaps.
In this paper we examine the consequences of structural transformation and urbanization for
rural-urban inequality by focusing on the experience of China and India since the 1980s. These
two countries are particularly relevant for studying the process of structural transformation and its
consequencesfortworeasons. First,alongthelinesofthestylizedfactsofdevelopmentdocumentedin
Kuznets(1966),theperiodofrapideconomicgrowthsincethe1980sinChinaandIndiahasalsobeen
markedbyasigni(cid:133)cantstructuraltransformationofthesetwoeconomieswithfactorsandproduction
moving away from agriculture towards manufacturing and services. This period and process has also
been accompanied by rising urbanization in both countries. Second, structural transformation has
potentially large distributional consequences because the contracting agricultural sector is typically
rural while the expanding sectors are typically urban. As a result this process typically requires
workers to not just change sectors but often change occupations as well as geographic location.
Given that upwards of 1.5 billion people still reside in the rural China and India in 2011 (which
is almost a (cid:133)fth of the world population), the scale of the potential disruption and reallocation
unleashed by this process is massive. Hence, the ongoing structural transformation is likely to have
(cid:133)rst order implications for overall inequality in the two countries.
A striking data feature that we document is that rural-urban income disparities have evolved in
very di⁄erent ways in the two countries. Between 1988 and 2008 the mean wage gap between rural
and urban workers in China has widened by 6 percentage points with the median gap widening by
an even larger 21 percentage points. In contrast, the mean urban-rural wage gap in India between
1983 and 2010 has declined by 24 percentage points while the corresponding median wage gap has
contracted by a massive 46 percentage points. Somewhat puzzlingly, we also (cid:133)nd that individual
worker attributes little of the movements in wages.
Given the large residual wage convergence left unaccounted for by conventional covariates of
wages, the second part of the paper focuses on providing a structural explanation for it. The model
we develop examines the explanatory power of aggregate shocks in accounting for the unexplained
2
wage convergence. Our choices of model building blocks are dictated by the joint requirements of
accounting for both the ongoing structural transformation of the economy as well as the rural-urban
wage convergence. We develop a multi-sector model with non-homothetic preferences along with
two aggregate shocks (cid:150)sectoral productivity growth and urbanization. The model predicts that
productivity growth should induce a structural transformation of the economy from agriculture to
non-agriculture but also cause urban wages to rise relative to rural wages, thereby widening the
urban-rural wage gap. Urbanization however causes the urban-rural wage gap to fall since it raises
the relative share of urban labor.
We calibrate the model to China and India separately and examine its quantitative predictions
for both structural transformation and wage gaps by feeding in the actually observed changes in
sectoral productivity and urbanization during the sample period. For India these two aggregate
shocks account for both the structural transformation and the movements in the wage gaps. For
China the model accounts for the observed structural transformation but also predicts a fall in the
urban-rural wage gap in contrast to the widening gap in the data. As a diagnostic tool for the
perverse wage outcome in China, we compute the wedges between the observed sectoral wages in the
data and the sectoral marginal products of labor and (cid:133)nd that the urban labor wedge grew about
12 percent faster than the rural labor wedge during this period. This suggests to us that the urban
labor market may have been the source of increasing distortion over this period.
Our mechanism for generating structural change relies on lower income elasticity of demand for
agriculturalgoodsduetothenon-homotheticityinpreferencesintroducedbytheminimumconsump-
tion need for the agricultural good and (cid:133)xed home production in the non-agricultural sector. This is
a demand-side e⁄ect generated by changing incomes. There is also a supply-side mechanism that has
been proposed in the literature (dating back to Baumol (1967)) which relies on di⁄erential sectoral
productivity growth. In particular, Ngai and Pissarides (2007) use a multi-sector model to show that
as long as the elasticity of substitution between (cid:133)nal goods is less than unity, over time factors would
move to the sector with the lowest productivity growth. In the Indian case, this mechanism leads
to a counterfactual implication. As we show, productivity growth in non-agriculture was faster than
in agriculture. Hence, the Ngai and Pissarides (2007) mechanism would imply that factors should
have migrated to the agricultural sector over time while the data shows the opposite. One could
get around this by assuming that the elasticity of substitution between (cid:133)nal goods is greater than
unity. However, given the lack of precise estimates on this elasticity, it seems heroic to put the entire
onus of the explanation on the con(cid:133)guration of a poorly measured parameter. Consequently, we shut
down this channel by assuming that the elasticity of substitution between (cid:133)nal goods is unity. This
also implies that the reasons for structural transformation in the model are the non-homotheticity
in preferences and increases in relative urban labor supply due to urbanization.1
1See Laitner (2000), Kongsamut, Rebelo, and Xie (2001) and Gollin, Parente, and Rogerson (2002) for a formal-
ization of the non-homothetic preference mechanism. Our work is also related to the factor deepening channel for
structuraltransformation formalized in Acemoglu and Guerrieri(2008). An overview ofthisliteraturecan befound in
3
Our supply-side channel is complementary to the skill acquisition cost mechanism proposed by
Caselli and Coleman (2001) in their study of regional convergence between the North and South of
the USA. Like our urbanization shock, in their model a fall in the cost of acquiring skills to work in
the non-agricultural sector induces a fall in farm labor supply and leads to an increase in farm wages
and relative prices.
Our focus on rural-urban gaps probably is closest in spirit to the work of Young (2012) who has
examined the rural-urban consumption expenditure gaps in 65 countries. Like us, he (cid:133)nds that only
a small fraction of the rural-urban inequality can be accounted for by individual characteristics, such
as education di⁄erences. He attributes the remaining gaps to competitive sorting of workers to rural
and urban areas based on their unobserved skills.2 Our work is also related to an empirical literature
studying rural-urban gaps in di⁄erent countries (see, for instance, Nguyen, Albrecht, Vroman, and
Westbrook(2007)forVietnam,WuandPerlo⁄(2005)andQuandZhao(2008)forChinaandothers).
These papers generally employ household survey data and relate changes in urban-rural inequality
to individual and household characteristics.
The rest of the paper is organized as follows: the next section presents the data and the main
resultsonchangesintherural-urbangapsaswellastheanalysisoftheextenttowhichthesechanges
were due to changes in individual characteristics of workers and their migration decisions. Section 3
presents our model and examines the role of aggregate shocks in explaining the patterns. The last
section contains concluding thoughts.
2 Empirical results
Our primary data source for China is the Chinese Household Income Project (CHIP). We use (cid:133)ve
rounds of the CHIP (1988, 1992, 1995, 2002 and 2008). Since our interest is in determining the
trends in wages and determinants of wages such as education and occupation, we choose to restrict
the sample to individuals in the working age group 16-65 who are identi(cid:133)ed as working and who
report working at least 1900 hours per year. These restrictions leave us with 47,000 to 83,000
individuals per survey round. The data for India comes from successive rounds of the Employment
& Unemployment surveys of the National Sample Survey (NSS) of households in India. The survey
roundsthatweincludeinthestudyare1983, 1987-88, 1993-94, 1999-2000, 2004-05, and2009-10. We
restrict the sample to individuals in the working age group 16-65, who are working full time (de(cid:133)ned
as those who worked at least 2.5 days in the week prior to being sampled), who are not enrolled in
any educational institution, and for whom we have both education and occupation information. We
Herrendorf, Rogerson, and Valentinyi (2013a).
2Young(cid:146)s explanation based on selection is complementary to Lagakos and Waugh (2012). Our (cid:133)nding of unex-
plained changes in rural-urban wage gaps over time also (cid:133)nds an echo in the work of Gollin, Lagakos, and Waugh
(2012) who (cid:133)nd large and unexplained di⁄erences in value-added per worker in agriculture relative to non-agriculture
in developing countries.
4
further restrict the sample to individuals who belong to male-led households.3 These restrictions
leave us with, on average, 140,000 to 180,000 individuals per survey round. Details on our data are
provided in Appendix A.1.
To obtain a measure of labor income we need wages and the occupation distribution of the labor
force. For China, we use annual wage income which are de(cid:135)ated using state-level CPI de(cid:135)ators
that are available separately for rural and urban sectors. For India we measure wages as the daily
wage/salaried income received for the work done by respondents during the previous week (relative
to the survey week), if the reported occupation during that week is the same as worker(cid:146)s usual
occupation (one year reference).4 Wages can be paid in cash or kind, where the latter are evaluated
at current retail prices. We convert wages into real terms using state-level poverty lines that di⁄er
for rural and urban sectors.5 We express all wages in 1983 rural Maharashtra poverty lines.6
WestartbycomputingthewagegapsbetweenurbanandruralworkersinChinaandIndia. Panel
(a) of Figure 1 shows the mean and median gaps for China while Panel (b) shows the corresponding
gaps for India. The panels present a striking contrast: both the mean and the median urban-rural
wage gaps widened in China between 1988 and 2008 while they narrowed in India between 1983 and
2010.
WeshouldnotethatwhiletheNSSdataforIndiaisanationallyrepresentativesample, theCHIP
data for China is not nationally representative as it samples households from a subset of provinces
only in each round. As a result one may be concerned about the robustness of the wage patterns
documented in Figure 1. To examine robustness, Figure 2 plots the urban-rural wage gaps computed
from two other sources as well as by alternative restrictions on the CHIP sample. All four plots
reveal the same pattern of widening wage gaps between urban and rural Chinese workers over the
past three decades.
One concern about the aggregate pattern of widening wage gaps in China and contracting gaps
in India is that they may be driven by some outlier states or regions. To check the robustness of
the patterns, Panel (a) of Figure 3 plots scatter of the urban-rural wage gaps across provinces in
China for 1990 and 2008 while Panel (b) plots the same scatter of wage gaps for states in India for
3This avoids households with special conditions since male-led households are the norm in India.
4This allows us to reduce the e⁄ects of seasonal changes in employment and occupations on wages.
5Using poverty lines that di⁄er between urban and rural areas may generate real wage convergence if urban prices
are growing faster than rural prices. This is indeed the case in India during our study period. However, only a small
fraction of the observed real wage convergence is driven by the price dynamics. In the online appendix we show that
nominal wages are converging slightly faster than real wages (except at the mean) during 1983-2010 period.
6In 2004-05 the Planning Commission of India changed the methodology for estimation of poverty lines. Among
otherchanges,theyswitchedfromanchoringthepovertylinestoacalorieintakenormtowardsconsumerexpenditures
more generally. This led to a change in the consumption basket underlying poverty lines calculations. To retain
comparability across rounds we convert the 2009-10 poverty lines obtained from the Planning Commission under the
newmethodologytotheoldbasketusinga2004-05adjustmentfactor. Thatfactorwasobtainedfromthepovertylines
under the old and new methodologies available for the 2004-05 survey year. As a test, we used the same adjustment
factortoobtaintheimplied"old"povertylinesforthe1993-94surveyroundforwhichthetwosetsofpovertylinesare
also available from the Planning Commission. We (cid:133)nd that the actual old poverty lines and the implied "old" poverty
lines are very similar, giving us con(cid:133)dence that our adjustment is valid.
5
Figure 1: The urban-rural wage gaps in China and India
(a) China (b) India
Notes: Panel (a) shows the urban-rural wage gaps for China for the 1988 and 2008 CHIP rounds. Panel
(b) shows the urban-rural wage gaps for India for the 1983 and 2009-10 NSS rounds. These are obtained
from a regression of log wages on a rural dummy and age controls.
Figure 2: The urban-rural wage gaps in China: robustness
CHIP: full time wages CHIP: p.c. family income
6 4
.1
5
.1 3
4
.1 2
3
.1
2 1
.1
1.1 0
1990 1995 2000 2005 2010 1990 1995 2000 2005 2010
mean gap median gap mean gap median gap
China statistical yearbook CHNS 1989 2006 mean wage
5 3
.3 .1
3 2.1
5 1
.2 .1
2 1
1990 1995 2000 2005 2010 1990 1995 2000 2005
Notes: The (cid:133)gure plots the urban-rural wage gap computed from di⁄erent sources and using di⁄erent
restrictions on the sample.
1983 and 2010. The key feature to note is most of the points for China lie above the 45 degree line
indicating larger gaps in 2008 relative to 1990. The corresponding scatter of points for Indian states
lie primarily below the 45 degree line indicating a narrowing of the wage gap between urban and
rural workers between 1983 and 2010.
6
Figure 3: The urban-rural wage gaps by province/state in China and India
5.2 AS
GS QH 2
GX
NX
XJ
SX
NM HANHUBSD FJHEHNEB 5.1 CHMNTR WB MBHR ML
80022 HLJ JLZJJX LN 0102 PBMPRZYJ HGUPJPAPKASKMP OR
1 KALN HR TN
JK
DL
5.1 5. GA
1.5 2 2.5 .5 1 1.5 2
1990 1983
mean income gap mean wage gap
(a) China (b) India
Notes: Panel (a) shows the urban-rural wage gaps for provinces in China for 1990 and 2008. Panel (b)
shows the urban-rural wage gaps for India for 1983 and 2009-10 NSS rounds.
What explains the observed patterns in the urban-rural wage gaps in the two countries? The
standard explanations focus on measured attributes in wages such as demographics, education, oc-
cupation, etc..How much of the wage convergence documented above is driven by a convergence
of measured covariates? We examine this using a procedure developed by DiNardo, Fortin, and
Lemieux (1996) (DFL from hereon) to decompose the di⁄erence in the observed wage distributions
of urban and rural labor within a sample round into two components (cid:150)the part that is explained
by di⁄erences in attributes and the part that is explained by di⁄erences in the wage structure of the
two groups. To obtain the explained part, for each set of attributes we construct a counterfactual
density for urban workers by assigning them the rural distribution of the attributes.7
We consider several sets of attributes. First, we evaluate the role of individual demographic
characteristicssuchasage,agesquared,andgender. Second,weaddeducationtothesetofattributes
and obtain the incremental contribution of education to the observed wage convergence. Lastly, we
evaluate the role played by di⁄erences in the occupation distribution for the urban-rural wage gaps.
Figure 4 presents our (cid:133)ndings for China for 1988 (panel (a)) and 2008 (panel (b)). The solid
line shows the actual urban-rural (log) wage gaps for the entire wage distribution, while the broken
lines show the gaps explained by di⁄erences in attributes of the two groups, where we introduced the
attributes sequentially. The plot for 1988 says that observed gaps in attributes explained very little
of the actual wage gap in that year. In 2008 by contrast, measured attributes, especially education,
explain almost 60 percent of the actual median gap with the explanatory power becoming smaller
7TheDFLmethodinvolves(cid:133)rstconstructingacounterfactualwagedensityfunctionforurbanindividualsbygiving
them the attributes of rural households. This is done by a suitable reweighting of the estimated wage density function
of urban households. We choose to do the reweighting this way to avoid a common support problem, i.e., there may
not be enough rural workers at the top end of the distribution to mimic the urban distribution. The counterfactual
urban wage density is then compared with the actual urban wage density to assess the contribution of the measured
attributes to the observed wage gap.
7
at the higher income percentiles. The two graphs combined suggest that a signi(cid:133)cant part of the
overall movement in the wage gaps was unaccounted for by measured covariates of wages.
Figure 4: Decomposition of urban-rural wage gaps in China for 1988 and 2008
Urban Rural wage gap, 1988 Urban Rural wage gap, 2008
8. 8.
6
. 6
.
4
.
4
.
2
.
2
.
0
2 0
.
0 20 40 60 80 100 0 20 40 60 80 100
percentile percentile
actual explained:demogr actual explained:demogr
explained:+edu explained:+occ explained:+edu explained:+occ
(a) China, 1988 (b) China, 2008
Notes: Eachpanelshowstheactuallogwagegapbetweenurbanandruralworkersforeachpercentile,and
the counterfactual percentile log wage gaps when urban workers are sequentially given rural attributes.
Three sets of attributes are considered: demographic (denoted by "demogr"), demographics plus educa-
tion ("edu"), and all of the above plus occupations ("occ"). The left panel shows the decomposition for
1988 while the right panel is for 2008.
Figure 5 plots the same graphs for India for 1988 (panel (a)) and 2009-2010 (panel (b)). Like
in China, demographic characteristics explain a small fraction of the urban-rural wage gap. In 1983
di⁄erences in education account for almost the entire wage gap at the bottom of the distribution,
while di⁄erences in occupation explain the wage gap for the upper 50 percent of the distribution.
Put di⁄erently, education and occupation choices can jointly account for almost the entire wage
gap distribution in India in 1983. In 2009-10 however, di⁄erences in education attainments between
urban and rural workers explain a large fraction of the gap at the top end of the distribution (70th
percentile and above). However, for those below the 70th percentile, covariates such as demographic
characteristics, education and occupation choices systematically over-predict the actual wage gaps.
This is particularly stark for the bottom 15 percent where the actual wage gap is negative while the
demographic characteristics, education endowments and di⁄erences in occupations predict that the
urban-rural gap should be positive 30 percent.
These results suggest that a large part of the observed convergence in wage di⁄erences cannot
be explained by standard covariates of wages. Hence, the wage structure of urban and rural workers
and changes therein during the sample period play an important role in our data.
8
Figure 5: Decomposition of urban-rural wage gaps in India for 1983 and 2009-10
Urban Rural wage gap, 1983 Urban Rural wage gap, 2009 10
8
.
7. 8.
)laruR456... )laruR4567...
(egaw23.. (egaw23..
nl )nab10. nl )nab10..
rU(egawnl321... rU(egawnl321...
0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100
percentile percentile
actual explained:demogr actual explained:demogr
explained:edu explained:occ explained:edu explained:occ
(a) India, 1983 (b) India, 2010
Notes: Eachpanelshowstheactuallogwagegapbetweenurbanandruralworkersforeachpercentile,and
the counterfactual percentile log wage gaps when urban workers are sequentially given rural attributes.
Three sets of attributes are considered: demographic (denoted by "demogr"), demographics plus educa-
tion ("edu"), and all of the above plus occupations ("occ"). The left panel shows the decomposition for
1988 while the right panel is for 2008.
3 The Role of Aggregate Shocks
The previous results suggest that a large part of the movement in the wage gap between rural and
urban India cannot be accounted for by convergence in the individual characteristics of the two
groups. What then explains the trends? One possibility is that aggregate developments during this
periodmayhaveplayedarole. Speci(cid:133)cally,theperiodsincethe1980swasmarkedbyasharpincrease
intheaggregategrowthrate,structuraltransformationandrapidurbanizationoftheeconomy. Could
these aggregate changes have contributed to the observed trends in rural-urban gaps in China and
India? In this section we examine this possibility by exploring their e⁄ects through the lens of a
structural model.
3.1 Key aggregate facts
Before presenting the model it is useful to summarize some key aggregate developments in China
and India over the last three decades. Our key aggregate facts relate to the structural composition
of employment and output, sectoral productivities, the urban share of the labor force, and relative
prices. We want the model to be consistent with these facts.
Note that below we present aggregate facts for industries rather than occupations. This is in-
nocuous since we will only distinguish between agriculture and non-agriculture based activities, and
because the vast majority of agricultural jobs are in the agriculture industry. This guarantees a tight
mapping between occupations and industries.
The ongoing process of structural transformation of the China and India can be seen through
9
Figures 6 and 7. Figure 6 shows employment shares in agriculture and non-agriculture for China
(panel (a)) and India (panel (b)). Figure 7 shows the distribution of output across the agriculture
and non-agriculture in the two economies. As is easy to see, agriculture has been releasing workers
in both countries, and its share of output has also been declining over time in both. The non-
agricultural sector, on the other hand, has expanded both as a share of employment and as a share
of output in both India and China. These are the textbook features of structural transformation.
Figure 6: Employment distribution
001 001
08 08
06 06
04 04
02 02
0 1988 2000 0 1983 1993 94 2004 05
1995 2008 1987 88 1999 00 2009 10
Agri Non agri Agri Non agri
(a) China: employment shares (b) India: employment shares
Notes: Panel (a) of this Figure presents the distribution of workforce across agricultural and non-
agriculturalsectorsforChinawhilepanel(b)presentstheemploymentdistributionacrossthetwosectors
for India.
Figure 7: Sectoral output distribution
001 001
08 08
06 06
04 04
0 0
2 2
0 1988 2000 0 1983 1993 94 2004 05
1995 2008 1987 88 1999 00 2009 10
Agri Non agri Agri Non agri
(a) China: output shares (b) India: output shares
Notes: Panel(a)ofthisFigurepresentsthedistributionofoutputacrossagriculturalandnon-agricultural
sectors in China.Chinea. Panel (b) presents same distribution for India.
The third key aggregate fact of interest is the behavior of sectoral labor productivities that
were underlying the process of structural transformation. Figure 8 presents labor productivity in
10