Population and Economic Change

The Effect of Urbanization on China’s Fertility

Zhen Guo • Zheng Wu • Christoph M. Schimmele •

Shuzhuo Li

Received: 15 September 2011 / Accepted: 24 January 2012 / Published online: 10 February 2012

� Springer Science+Business Media B.V. 2012

Abstract The relationship between urbanization and fertility decline is known to be inverse in developed countries. However, the nature of this relationship in

developing countries that already have relatively low fertilities is not well-under-

stood. This study aims to illustrate how much urbanization contributed to China’s

fertility decline between 1982 and 2008 and forecasts how much it can contribute to

future reductions in fertility. The study examines changes in the total fertility rate

(TFR) at both the national and provincial levels, given regional differences in the

urbanization rate. The results show that changes in rural fertility behavior accounted

for most of the decline in the national TFR between 1982 and 2008. This finding

suggests that official birth control policies were instrumental in curbing China’s

population growth. However, urbanization was responsible for about 22% of the

decrease in TFR during this period, and its effect was especially important during

the latter years (2001–2008). In most provinces, urbanization associated with a

decline in provincial-level fertility. The forecasts indicate that urbanization will

become the primary factor behind future declines in national fertility. Given the

negative effect of urbanization on the TFR, it is possible to relax the one-child

policy without having adverse implications for population growth.

Keywords Urbanization � Fertility � China

Z. Guo

School of Management, Xi’an Jiaotong University, Xi’an, China

Z. Wu (&) � C. M. Schimmele Department of Sociology, University of Victoria, 3800 Finnerty Road,

Victoria, BC V8W 3P5, Canada

e-mail: [email protected]

S. Li

Institute for Population and Development Studies, School of Public Policy and Administration,

Xi’an Jiaotong University, Xi’an, China

123

Popul Res Policy Rev (2012) 31:417–434

DOI 10.1007/s11113-012-9230-0

Introduction

China’s total fertility rate (TFR) declined from 2.78 in 1978 to 1.47 in 2008 (National

Bureau of Statistics of China 2009). The TFR decreased to sub-replacement levels

sometime during the early 1990s (Feeney and Yuan 1994). The pace of this decline is

remarkable considering that the TFR was over 5.0 until the 1970s (Gu 2007). This

change has been attributed primarily to the Chinese government’s efforts to curb

population growth, such as the one-child policy (Feeney and Wang 1993). These

birth control interventions certainly set China’s demographic transition apart from

other transitions to low-fertility. However, whether birth planning policies are the

primary reason for China’s low fertility is not uncontested (Cai 2010). At least, it

appears that, similar to the demographic transitions in Western countries, socioeco-

nomic forces have also contributed substantially to China’s transition (Poston and

Gu 1987).

In most countries, there is an inverse relationship between TFR and socioeco-

nomic development, with fertility declining as development progresses (Bongaarts

and Watkins 1996; Bryant 2007). This is demonstrated in the long-established TFR

differential between urban and rural areas (Jaffe 1942). Given that urbanites tend to

have/prefer fewer children than rural residents, the process of urbanization propels a

reduction of national TFR. Over two decades ago, Zeng and Vaupel (1989)

observed that this process would likely decrease future birth rates in China. At that

time, the urban–rural fertility differential remained quite large. In 1986, the urban

TFR was 1.96 compared to the rural TFR of 2.72. Since China was predominantly a

rural, agricultural society in the mid-1980s, Zeng and Vaupel anticipated that the

national TFR had much room to decline through rural-to-urban migration and the re-

classification of rural areas into urban areas. This would occur as former rural

residents voluntarily adopted the preference for fewer children that is prevalent

among urbanites or were compelled to have fewer children because of the stricter

enforcement of the one-child policy in urban areas.

At the time Zeng and Vaupel made this observation, China’s TFR was above the

replacement level and almost two-thirds of the population lived in rural areas. At

present, China’s TFR is 1.47 (see Fig. 1) and over 46% of the population resides in

urban areas (United Nations 2010). The proportional size of China’s urban

population is below the global average (50%) and far below the average (75%) for

developed countries. Hence, the potential for urban growth is large and it is

expected that 73% of the Chinese population will live in urban areas in 2050. What

is uncertain is how much urbanization can contribute to future reductions in China’s

TFR. In general, our knowledge is limited about the determinants of fertility

behavior in countries that are undergoing the process of development but have low

fertility (Bongaarts 2002). This leaves questions about the relationship between TFR

and urbanization in China, which cannot be considered a developed country, but has

achieved sub-replacement fertility.

In China, the fertility differential between rural and urban areas has narrowed

since 1978, but it is still large (see Fig. 2). In 2008, the TFR was 1.73 in rural areas

and 1.22 in urban areas. The rural–urban TFR differential has, moreover, remained

fairly stable since the early 1990s. If urban fertility behavior remains consistent, this

418 Z. Guo et al.

123

implies that urban expansion will propel further reductions in China’s fertility.

According to Bongaarts (2002), at high levels of development the relationship

between TFR and socioeconomic indicators is likely to be nonlinear, because it is

unreasonable to expect an indefinite decline in TFR as socioeconomic development

progresses. That is, although socioeconomic development corresponds to a

reduction in fertility, it cannot totally extinguish the desire for children. For China,

this relationship could become nonlinear at a comparatively lower stage of

development as China’s TFR is already among the lowest in the world. The

continuing urbanization of China appears to be inevitable, but it is likely that at

some point this process will no longer lead to further reductions in TFR.

The issue here is whether the relationship between TFR and urbanization is

weakening in the Chinese context. The demographic trends suggest this is the case.

0

0.1

0.2

0.3

0.4

0.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008

T F

R

TFR Urbanization rate

Year

D egree of U

rbanization

Fig. 1 Trends of TFRs and urbanization rates, 1978–2008, China. Source: National Bureau of Statistics of China (2009)

0.5

1

1.5

2

2.5

3

3.5

1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008

T F

R

Rural

Urban

Year

Fig. 2 Trends of TFRs in urban and rural areas in 1978–2008. Sources: The 1978–1987 data came from the 2/1,000 fertility and contraceptive use survey conducted in 1988; the 1988–1992 data came from the National Fertility Survey conducted in 1992; the 1993–2000 data came from the 2001 National Fertility and Reproductive Health Survey; and the 2001–2008 data came from the Annual Population Monitoring Surveys

The Effect of Urbanization on China’s Fertility 419

123

Figure 1 illustrates that urbanization corresponded with a large decline in TFR from

1978 and sometime in the mid-1990s. However, the TFR plateaued thereafter, even

though the urbanization rate kept increasing. This raises the question of how much

fertility changes (or can change) in response to urbanization in low-fertility regimes.

To address this question, this study uses decomposition models to assess the

contribution of urbanization to the decline in China’s TFR since 1978. The study

examines both the national and provincial levels because there are regional

differences in socioeconomic development and the enforcement of national birth

planning policies. In addition, the study simulates how much more urbanization can

be expected contribute to fertility decline, under several alternative scenarios of

urban expansion and birth planning reforms, forecasting TFR until 2030.

Background

A key debate in the literature regards the primary source of China’s fertility decline.

The debate is about how much socioeconomic factors have contributed to this

decline, given the Chinese government’s tight regulation of fertility behavior. The

predominant notion is that birth planning policies are the fundamental reason for

China’s demographic transition (Poston and Gu 1987). These interventions—which

reportedly have prevented over 300 million births since 1978 (Peng 2004)—have

led to doubts about whether the socioeconomic indicators that were instrumental to

fertility decline in Western countries are also good explanations for China’s

demographic transition. In China, the congruence of the timing of fertility change

with the implementation of birth control policies is clear evidence for the

importance of government intervention (Feeney and Wang 1993). For this reason,

the Chinese transition is considered to be a unique case among the countries that

have reached sub-replacement levels of fertility (Cai 2010).

Though best known is the controversial one-child policy, China’s efforts to

control population growth began well before this. The earliest interventions came as

a response to the Great Famine of 1959–1961. After this natural disaster, the

government began to set official targets for population growth and provide better

access to birth planning information (Wu et al. 2009). However, birth control did

not become a core aspect of economic planning until the 1970s (Sharping 2003).

This started with the Wan-Xi-Shao (later-longer-fewer) campaign, which promoted later marriage and childbirth, longer birth intervals, and fewer births (Liang and Lee

2006). Coinciding with the economic reforms, the Deng Xiaoping administration

implemented the one-child policy in 1978 to improve China’s prospects for

modernization and industrialization and to address concerns about foodgrain

shortages (Wu et al. 2009). The one-child per couple rule applies to around two-

thirds of Chinese couples, with most concessions to this rule applying to couples

residing in rural areas (Cai 2010).

Feeney and Wang (1993) suggest that over one-half of China’s fertility decline is

attributable to state intervention. There is no doubt that government policies hastened

the ‘‘diffusion’’ of low fertility throughout China. In 1975, the TFR was lower than

what could be expected from the level of development at that time, and reflected the

success of the Wan-Xi-Shao campaign (Cai 2010). But state intervention, and the

420 Z. Guo et al.

123

one-child policy in particular, is not the sole reason for the decline in fertility. The

persistence of sub-regional variation in fertility after the intensification of state

interventions appears to parallel sub-regional differences in socioeconomic devel-

opment (Tien 1984). Cai (2010) observes that China’s fertility in 2005 fell within a

range that could be expected from its level of development. In addition, the TFR

remained above replacement levels during the 1980s, when the one-child policy was

enforced with fewer exceptions than later. Cai concludes that socioeconomic

development, in conjunction with state intervention, generated an ideational shift

toward a preference for smaller families. For the one-child policy to be the sole or

decisive factor, the fertility behavior of Chinese couples would need to have been

radically different from the fertility behavior of couples in other countries.

To some extent, China’s path to low fertility supports the assumptions of

demographic transition theory (DTT). A central theme of DTT is that the shift from

rural (agricultural) to urban (industrial) life initiates a change in the economics of

childbearing (Kirk 1996). According to Notestein’s (1953) classic argument, fertility

is high in agrarian societies as insurance against high mortality and because children

were an important source of agricultural labor. Modernization first leads to a reduction

in mortality, which decreases the need for high fertility to insure population survival

(Bongaarts and Watkins 1996). The transition to industrial economies (and urban

environments) also decreases the economic contributions of children, whereas the

costs of their upbringing and education increase. Though no two transitions are alike,

it remains plausible that modernization is responsible for decreasing the need and

incentives for large families in numerous societies (Kirk 1996).

This theory of fertility change has been criticized for over-emphasizing the role

of economic motivation (Hirschman 1994). To be sure, the precise reasons for the

relationship between TFR and socioeconomic development are difficult to ascertain,

and are surely irreducible to economic factors. Even though DTT offers an

incomplete explanation of fertility change, this does not undermine the empirical

relationship between TFR and levels of socioeconomic development (Bryant 2007;

Cai 2010). The main criticism of DTT is not the relationship between modernization

and fertility per se, but the mechanisms that constitute this relationship (Bongaarts

and Watkins 1996). The criticism of DTT also focuses largely on the role of

socioeconomic indicators in the onset and early phase of the transition to low

fertility. However, Bongaarts (2002) observes that fertility behavior is more

consistent with DTT at later stages of the transition, which is our concern.

Of course, the relationship between fertility and socioeconomic development

cannot be reduced to rational decisions about the costs/benefits of children

(Hirschman 1994). However, DTT does not preclude other causal variables and

indeed acknowledges the importance of ideational factors. Notestein (1953)

observed that it is ‘‘impossible to be precise’’ about the mechanisms that drive

fertility change in modern societies, and he indicated that economic factors cannot

provide a sufficient explanation. He remarked that the anonymity of urban life

weakened social control over fertility behavior and modernization created more

opportunities for women outside the domestic sphere. Urbanization is a proxy for

changes in social norms and gender roles, which, together with economic forces,

generate a preference for smaller families.

The Effect of Urbanization on China’s Fertility 421

123

The economics of children and ideational preferences for smaller families are

important components of the relationship between fertility decline and socioeco-

nomic development in China (Cai 2010). However, rural–urban differences in the

enforcement of the one-child policy suggest that an increasing proportion of urban

residents will lead to an inevitable decline in national fertility, unless China reforms

the policy. The one-child rule is strictly enforced in all urban areas in China and

throughout 6 provinces (Gu et al. 2007). There are some exceptions for couples that

have agricultural household registration status. In 19 provinces rural couples are

allowed a second child if their first child is a girl and in another 5 provinces all rural

couples are permitted two children. The urban population remained stable until

1978, but the relaxation of official restrictions on rural-to-urban migration and the

reclassification of rural areas into urban areas have fueled the proliferation of the

urban population (Zeng and Vaupel 1989). This process is exposing a growing

number of Chinese to urban values and subjecting them to the one-child rule.

Methods

The data for the country-level TFR and the proportion of urban females of

reproductive age come from the 1982 Census, the 1990 Census, and the 2001 and

2008 one per thousand population surveys conducted by the National Bureau of

Statistics of China (NBS 2009). The estimates for provincial-level fertility are

drawn from an NBS and East–West Center (2007) report. A decomposition

approach is used to model the effects of urbanization on fertility change. Following

Das Gupta (1991), the analysis decomposes TFRasfr into three components to

estimate the separate effects of changes in urban fertility, rural fertility, and

urbanization on TFR. For the reader’s convenience, the mathematical expression is

recapitulated as below.

TFR can be formulated as TFRasfr ¼ 5 P

x Fx, where Fx is the age-specific birth rate for the 5-year age group starting at age x. Fx can be expressed as a weighted sum of urban-age-specific birth rate(Fx,u) and rural-age-specific birth rate (Fx,r), where the weights kx,r and kx,u are the proportion of women in age group x to x?5 residing in rural and urban areas, respectively (here we have kx,r ? kx,u = 1, Dkx,u = -Dkx,r). This leads to the reformulation of TFRasfr,

TFRasfr ¼ 5RxFx ¼ 5Rx Fx;rkx;r þ Fx;ukx;u � �

ð1Þ It follows that the change in the TFRasfr is,

DTFRasfr ¼ 5Rx Fx;u � Fx;r � �

Dkx;u þ 5Rxkx;rDFx;r þ 5RFx;u DFx;u ð2Þ

where the symbol D denotes change, and Fx;r, Fx;u, kx;r and kx;u are average values over the period. The first of the three principal terms on the right hand side of

Eq. 2 denotes the contribution to change in TFR from changes of the age-specific

proportion of urban females within the total female population at reproductive

age. The second term denotes the contribution from changes in age-specific rural

fertility. The third term denotes the contribution from changes in age-specific of

urban fertility.

422 Z. Guo et al.

123

To demonstrate the results of this decomposition exercise, we begin with

the scenario where (i) the rural fertility is always higher than urban fertility in any

age-group, (Fx,r [ Fx,u), (ii) all components have no changes during the period (Dkx,u = DFx,r = DFx,u = 0) This situation is illustrated in Fig. 3a, in which all TFRs are constant during the period.

Suppose now that, under the same assumptions (i and ii), we now allow the proportion

of urban females to increase at each reproductive age (Dkx,u [ 0). Equation 2 is simplified: DTFRasfr ¼ 5Rx Fx;u � Fx;r

� � Dkx;u where Fx;u � Fx;r\0 according to

assumption (i) and DTFRasfr \ 0. As shown Fig. 3b, this change drives down the national fertility despite that both urban and rural fertility remains unchanged.

Furthermore, if there is a positive change in both urban and rural age-specific birth rate

(DFx,u [ 0 and DFx,r [ 0) in the proportion of urban females (Dkx,u [ 0), the first term of Eq. 2 becomes negative but the second and third terms turn positive, such that the

change of national TFR, as the sum of the three terms, can be unchanged

(DTFRasfr = 0). This scenario is demonstrated in Fig. 3c. In short, this illustration demonstrates that the trends of national fertility, urban fertility and rural fertility may

not be in the same direction when we take into account the role of urbanization.

National TFR, 1982–2008

Table 1 presents the change in age-specific national TFR, which is decomposed into

three components. The first component represents changes in rural fertility, the

TFR

Time

TFR(rural)

TFR(nation)

TFR(urban)

TFR

Time

TFR(rural)

TFR(nation)

TFR

Time

TFR(rural)

TFR(nation)

(a) (b)

(c)

TFR(urban)

TFR(urban)

Fig. 3 Illustrations of urbanization effects on fertility

The Effect of Urbanization on China’s Fertility 423

123

second component represents changes in urban fertility, and the third component

represents the influence of urbanization, i.e., changes in the proportion of urban

females aged 15–49 years. The results show a 1.15 decrease in China’s TFR

between 1982 and 2008. The change in rural fertility behavior contributed to 0.83 of

this decrease and the change in urban fertility behavior contributed to 0.07 of this

decrease. The change of urbanization resulted in a 0.25 decrease in the national

TFR, which represents about 22% of the total reduction in TFR from 1982 to 2008.

The change in rural fertility behavior accounted for the largest amount (72%) of the

decline in TFR during this period.

From 1982 to 1990, both the changes in rural fertility behavior and urbanization

led to a reduction in national TFR. During this period, there was an increase in

urban births, and thus urban fertility behavior had a positive impact on national

TFR. The reduction of national TFR through rural fertility behavior and

urbanization likely reflect the impact of the one-child policy. The results suggest

a tightening up of the one-child policy in rural areas, such as preventing 3rd and

higher order births. The effect of urbanization is presumably a result of a greater

number of people becoming adherents to the strict one-child rule through permanent

migration or the reclassification of rural areas into urban areas. The impact of urban

fertility behavior is not that surprising. Urban fertility has been considerably lower

than rural fertility since the 1960s and it reached the sub-replacement level in the

early 1970s (Zeng and Vaupel 1989). Given that the urban TFR was 1.4 in 1981

(Fig. 2), it is unreasonable to anticipate that it could decline much more.

Changes in rural fertility behavior, urban fertility behavior, and levels of

urbanization all contributed to the reduction in national TFR from 2.30 in 1990 to

1.39 in 2001. The change in rural fertility behavior had the greatest effect,

accounting for 69% of the decline in national TFR. The effect of urban fertility

behavior accounted for 20% of the decline in TFR and urbanization accounted for

the remaining 11%. From 2001 to 2008, the national TFR increased from 1.39 to

1.47. This was a result of growth in both rural and urban fertilities. However, the

effect of urbanization on national TFR growth was negative. The rebound of

national TFR demonstrates the challenge of reducing TFR in low-fertility regimes.

But these results also show that urbanization remains a source of declines in TFR in

developing countries with low fertility.

Table 1 demonstrates that urbanization was primary reason for the decline in

China’s TFR between 2001 and 2008. To illustrate the independent effect of

urbanization on fertility change, we compared the national TFR with and without

Table 1 Decomposition of the Changes in TFR in China: 1982–2008

Period TFR (per 1,000) Absolute change (per 1,000)

Start End Change Rural Urban Urbanization

1982–1990 2,620 2,300 -320 -310 60 -70

1990–2001 2,300 1,390 -910 -630 -180 -100

2001–2008 1,390 1,470 80 110 50 -80

1982–2008 2,620 1,470 -1150 -830 -70 -250

424 Z. Guo et al.

123

the effect of urbanization. In Fig. 4, the dotted line represents what the national TFR

would be without urbanization (counter-factual test). This figure confirms the

importance of urbanization to the decline in China’s fertility. Without urbanization,

China’s TFR would be higher than it actually is.

Province-Level TFR in 2000 and 2005

In this section, we present the decomposition of changes in TFR for 30 of 31

provinces and municipalities in mainland China. The region of Tibet is excluded

because the sample size of birth numbers is too small to permit an accurate analysis.

In China, socioeconomic development has been uneven and there are disparities

between the provinces (Peng 2011). The national results presented above could,

therefore, provide an incomplete picture of the relationship between TFR and

urbanization. China’s provinces and municipalities fall under four levels of

urbanization (Fu et al. 2009). The first level includes municipalities such as

Shanghai, Beijing, and Tianjin, which are China’s economic powerhouses and have

the highest national levels of urbanization. The second are provinces at a medium

level of urbanization, including Heilongjiang, Jilin, and Liaoning. The third level

consists of nine provinces with low levels of urbanization: Guangdong, Jiangsu,

Shandong, Hubei, Shanxi, Qinghai, Xinjiang, Hainan, and Ningxia. The remaining

provinces have very low levels of urbanization.

Figure 5 plots the province-level TFRs according to degree of urbanization. This

figure indicates that there is, in general, an inverse relationship between TFR and

urbanization. In accordance, the most urbanized provinces also had the lowest TFRs

in 2000 and 2005. However, there are incidences where low levels of urbanization

are associated with high levels of fertility. Table 2 provides additional evidence for

1

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

2001 2002 2003 2004 2005 2006 2007 2008

T F

R

Year

Nation

Counter-factual

Urban

Rural

Fig. 4 Counter-factual test on TFRs from 2001 to 2008. Source: National Bureau of Statistics of China (2009)

The Effect of Urbanization on China’s Fertility 425

123

this relationship. Between 2000 and 2005, both changes in urban fertility behaviors

and levels of urbanization contributed to a decrease in the national TFR, but these

effects were somewhat offset because of an increase in the rural TFR across China.

During this time, the TFR declined in 23 provinces/municipalities. The greatest

decreases occurred in the four large metropolitan municipalities, Beijing, Tianjin,

Shanghai, and Chongqing. The majority of reduction in these provincial TFRs is

attributable to changes in urban fertility behavior and the expansion of the urban

population. Other provinces also experienced large reductions in their TFRs. These

include three coastal provinces, Liaoning in the north and Guangdong and Hainan in

the south, and three inland provinces, Shanxi, Henan, and Jiangxi.

However, the relationship between TFR and socioeconomic development is not

entirely consistent across China. Several lesser developed provinces (Guizhou,

Yunnan, Qinghai, and Xinjiang) also experienced large declines in their TFRs. The

declines in these provinces were largely a result of changes in fertility behavior in rural

areas. In six inland provinces (Guangxi, Sichuan, Hubei, Jiangsu, Hebei, and Anhui)

the TFR increased. In some of these provinces, the relationship between TFR and

urbanization does not appear to be as robust as it is elsewhere, but this is generally

because high fertility in rural areas offset the effect of urbanization. Shandong is the

only coastal province that experienced a large increase in its TFR. That said, Fig. 6

illustrates that, between 2000 and 2005, increases in levels of urbanization associated

with a decline in the TFRs in all provinces except for Jilin, Shanghai, and Xinjiang.

The decompositions presented in Table 2 suggest that changes in urban fertility

behavior in the most urbanized provinces accounted for most of reduction in TFR

observed among them. These provinces (and municipalities) are Beijing, Tianjin,

Shanghai, Liaoning, Jilin, Heilongjiang, Zhejiang, and Guangdong. In these

provinces, an average of 65% of women aged 15–49 reside in urban areas,

compared to the national average of 41%. The declines in the number of urban

500

1000

1500

2000

2500

0.2 0.4 0.6 0.8 1.0

T F

R (

pe r

1, 00

0)

Degree of urbanization

2000 2005 2000 2005

Fig. 5 Trends of urbanization rates and TFRs in 2000 and 2005. Sources: ‘‘Fertility estimates for provinces of China’’ (National Bureau of Statistics and the East–West Center 2007) and The report of China’s 2005 national 1% population survey. Beijing: National Bureau of Statistics of China 2006

426 Z. Guo et al.

123

births in these provinces represented an important source of the decrease in the

national TFR. While changes in rural fertility behavior contributed much to decline

in the national TFR from 1982 to 1990 and also from 1990 to 2001 (Table 1), this

effect seems to have ebbed in recent years. Between 2000 and 2005, rural fertility

had a positive effect on the national TFR, even though this effect was offset because

of decreases related to urban fertility behavior and urbanization.

Table 2 Decomposition of the changes in TFR: Chinese Provinces, 2000–2005

Absolute change in TFR (per 1,000) Relative change in TFR (percent)

Total Rural Urban Urbanization Rural Urban Urbanization

China -74 23 -51 -45 1.6 -3.7 -3.2

Beijing -202 2 -189 -16 0.2 -21.7 -1.8

Tianjin -177 -3 -146 -27 -0.3 -14.8 -2.7

Hebei 101 142 -6 -36 9.7 -0.4 -2.4

Shanxi -173 -85 -72 -15 -5.2 -4.4 -0.9

Neimenggu -27 -4 3 -26 -0.3 0.3 -2.2

Liaoning -104 21 -98 -28 1.9 -9 -2.5

Jilin -13 54 -69 2 5.4 -7 0.2

Heilongjiang -12 52 -63 -1 5.2 -6.2 -0.1

Shanghai -389 -20 -371 2 -1.9 -34.8 0.2

Jiangsu 52 152 -49 -52 13.6 -4.4 -4.6

Zhejiang -69 37 -79 -27 2.8 -6 -2.1

Anhui 301 323 61 -83 21.8 4.1 -5.6

Fujian -29 104 -80 -53 8.5 -6.5 -4.4

Jiangxi -120 -70 18 -68 -3.9 1 -3.8

Shandong 253 204 84 -35 16.1 6.6 -2.7

Henan -301 -221 -44 -36 -13.9 -2.8 -2.2

Hubei 119 104 41 -25 8.3 3.3 -2

Hunan -16 13 11 -40 0.9 0.7 -2.7

Guangdong -421 -78 -313 -30 -5.4 -22 -2.1

Guangxi 43 88 2 -47 5 0.1 -2.7

Hainan -226 -56 -102 -68 -3.1 -5.5 -3.7

Chongqin -239 -36 -92 -112 -2.5 -6.4 -7.8

Sichuan 36 92 -30 -27 6.3 -2 -1.8

Guizhou -450 -367 -50 -32 -15.4 -2.1 -1.3

Yunnan -307 -240 -19 -49 -11.8 -0.9 -2.4

Shaanxi -85 -11 -51 -23 -0.9 -4 -1.8

Gansu -29 -7 -13 -10 -0.5 -0.9 -0.6

Qinghai -435 -275 -138 -22 -15.1 -7.5 -1.2

Ningxia -19 30 -4 -45 1.7 -0.2 -2.6

Xinjiang -156 -109 -61 14 -6.3 -3.5 0.8

Sources: Fertility estimates for Provinces of China National Bureau of Statistics and the East–West Center (2007) and The report of China’s 2005 national 1% sample survey. Beijing: National Bureau of

Statistics of China 2006

The Effect of Urbanization on China’s Fertility 427

123

In 12 provinces, change in rural fertility behavior was instrumental in propelling

either the growth or the reduction of provincial-level fertility between 2000 and

2005. These 12 provinces can be classified has having comparatively small urban

populations. In seven of these provinces (Hebei, Jiangsu, Anhui, Shandong, Hubei,

Guangxi, and Sichuan) the provincial-level TFR increased because of increases of

fertility in rural areas. In several of these provinces the TFR increased despite a

decrease in urban fertility and a negative effect of urbanization. Moreover,

urbanization had a negative effect on the TFR in each of these provinces, and

fertility in urban areas increased in only in Anhui and Guangxi. In some provinces,

such as Henan, Guizhou, and Yunnan, the reduction in their TFRs was mainly a

result of declines of fertility in rural areas.

Future Effects of Urbanization

The evidence presented above suggests that urbanization is an important factor in

the reduction of China’s TFR. The question that remains is whether urbanization

will have a negative effect on China’s fertility in the future. To address this

question, we forecasted China’s fertility from 2010 to 2030, using six scenarios

based on three assumptions about urban growth and two assumptions about

differences in rural and urban fertilities. Under our low-growth assumption, 62% of

the population will be urban in 2030. In the medium-growth assumption, the

proportion of the urban population will be 67% in 2030. In the high-growth

assumption, the urban population will account for 84% of the general population in

-8 -7 -6 -5 -4 -3 -2 -1 0 1

Chongqin Anhui Jiangsu Fujian Jiangxi Hainan Hunan Guangxi Shandong Tianjin Ningxia Liaoning Yunnan Hebei Henan Neimenggu Guangdong Zhejiang Hubei Sichuan Beijing Shaanxi Guizhou Qinghai Shanxi Gansu Heilongjiang Jilin Shanghai Xinjiang

Percent changeFig. 6 Effect of change in the urbanization rate on TFR in 2000–2005, Chinese Provinces

428 Z. Guo et al.

123

2030. The figures for the medium-growth scenario best accord with official

estimates of future urbanization (Pan and Wei 2010). Because predicted data are not

age-specific based, a simplified version of the decomposition equation is introduced

and presented in Appendix A.

We considered these three assumptions about urban growth under two different

assumptions about future differences in rural and urban fertilities. First, we used a

time series model to project the stochastic pattern of rural and urban fertilities.

Details about the stochastic model are presented in Appendix B. In this model, rural

TFR is stable at 1.6 and urban TFR is stable at around 1.1, for a fairly persistent

difference of 0.5 between them. Second, we used a model of the rural–urban TFR

differential that assumes that the birth planning policy has been relaxed to a two-

child rule for all couples. Under this assumption, the rural TFR would be 1.88 and

the urban TFR 1.5 in and after 2010 (see Zheng 2004). While the second assumption

suggests a narrowing gap of rural and urban TFRs (0.38), it is unreasonable to

expect rural and urban fertility behaviors will converge in the next 20 years, even if

the one-child policy is relaxed in urban areas.

Table 3 presents the estimated TFRs under these six scenarios of urbanization

and differences in rural and urban fertility. In all six scenarios, urbanization is

projected to be the primary factor behind fertility change from 2010 to 2030, and the

national TFR will remain at sub-replacement levels. The stochastic projections

result in little variation in rural and urban fertilities during this time. Changes in

rural and urban fertility behaviors are projected to have small negative effects on

China’s TFR under the present birth planning policy. Under the ‘‘relaxed policy’’

assumption, rural and urban fertility do not affect the national TFR. This implies

that the projected decline in TFR will occur entirely through urbanization. Under

medium-growth (the expected level of urbanization), the national TFR will decrease

from 1.44 in 2010 to 1.25 in 2030 under the stochastic assumption and from 1.7 to

1.6 under the ‘‘relaxed policy’’ assumption.

Conclusions

China has experienced rapid urbanization since 1978 and the urban population is

projected to continue growing for several more decades. As noted above, there is an

Table 3 Decomposition of the predicted TFRs: 2010–2030

Urbanization

development

Fertility

assumption

TFR (per 1,000) Absolute change (per 1,000)

Start End Change Rural Urban Urbanization

High growth ‘‘Stochastic’’ 1,440 1,200 -240 -20 -30 -190

‘‘Relaxed’’ 1,700 1,560 -140 0 0 -140

Medium growth ‘‘Stochastic’’ 1,440 1,250 -190 -20 -30 -140

‘‘Relaxed’’ 1,700 1,600 -100 0 0 -100

Low growth ‘‘Stochastic’’ 1,440 1,310 -130 -10 -40 -80

‘‘Relaxed’’ 1,700 1,640 -60 0 0 -60

The Effect of Urbanization on China’s Fertility 429

123

inverse relationship between TFR and urbanization. This study examined the effects

of urbanization on fertility change in China between 1978 and 2008, and projected

how much more urbanization can be expected to contribute to fertility change

between 2010 and 2030. This study decomposed China’s present and future TFR

into three components to estimate the separate effects of changes: the effect of

change in rural fertility behavior, the effect of change in urban fertility behavior,

and the effect of urbanization. The study assumed that regional differences in levels

of urbanization could influence the relationship between national TFR and

urbanization. Hence, the analysis includes findings for the decomposed effects on

the national and provincial-level TFRs for 2000–2005.

The study offers three major conclusions about past and future fertility trends.

First, the change in rural fertility behavior accounted for most of the decline in the

national TFR from 1982 to 2008. The national TFR declined from 2.62 to 1.47

during this period. The reduction in rural fertility was responsible for 72% of this

decline. This finding suggests that the one-child policy was the primary instrument

of China’s achievement of sub-replacement fertility. Between 2000 and 2005,

several less developed provinces (e.g., Guizhou, Yunnan, Xinjiang) experienced

large declines in their TFRs largely because of reductions in the number of rural

births. It is possible that some of these declines in rural fertility is related to other

aspects of socioeconomic development, such as improvements in the educational

attainment of rural residents or decreases in need for agricultural labor, but the one-

child policy is likely the main factor for this change. In seven provinces, however,

an increase of fertility in rural areas was the driving factor for increases in province-

level TFR, which could reflect local variation in the enforcement of the one-child

policy. As the majority of Chinese (54%) still live in rural areas, it is unsurprising

that this population remains the vanguard of China’s fertility transition.

Second, the contribution of urbanization to the decline of China’s TFR between

1982 and 2008 was modest in comparison to the large effect that decreases in rural

fertility had. However, urbanization was indeed an important factor and it had a

negative effect on the national TFR in each of the periods observed (1982–1990,

1990–2001, 2001–2008, and 1982–2008). About 22% of the reduction in the

national TFR between 1982 and 2008 is related to the process of urbanization.

Moreover, the findings suggest that urbanization has recently become the principal

source for curbing population growth. From 2001 to 2008, urbanization had a

negative effect on the national TFR, but increases in rural and urban births offset

this effect. In all but three provinces, urbanization was associated with a decline in

province-level TFRs between 2000 and 2005. The three exceptions (Jilin, Shanghai,

and Xinjiang) had relatively low rates of urbanization during this period, thus the

impact of urbanization on TFR in these areas was also minimal. Low rates of

urbanization and possible measurement errors in TFRs may explain the unexpected

relationship between TFR and urbanization among them. In contrast, in provinces

with high rates of urbanization and large rural–urban fertility differentials, the effect

of urbanization on province-level TFR is quite pronounced.

Given the short period of observation for changes in province-level TFRs

(5 years), it is possible that the findings presented here do not reflect the full effect

of urbanization. The intent here is to disentangle the effect of urbanization from the

430 Z. Guo et al.

123

effect of urban fertility behavior. To some extent, the change in urban fertility

behavior is likely a ‘‘lagged’’ effect of urbanization. That is, the effects of rural-to-

urban migration and the reclassification of rural areas into urban areas on urban

fertility are not immediate. Rather, these new urbanities gradually adopt urban

fertility behaviors and are exempt from the strict one-child rule in the short-term.

This effect is difficult to decompose because of data limitations, but it suggests that

a portion of the decreases in the national and province-level TFRs related to changes

in urban fertility behavior represent an unobserved effect of urbanization.

Finally, the findings suggest that urbanization will take over as the main engine

of fertility decline from 2010 to 2030. This is evident from recent trends. While the

national TFR increased from 1.39 to 1.47 from 2001 to 2008, this change would

have been larger without the negative effect of urbanization. After several decades

of birth planning, it appears that the one-child policy is reaching the limits of what it

can accomplish. Our projections indicate that changes in rural and urban fertility

behaviors have small effects on the TFR under the current policy. In general, it is

becoming increasingly difficult to decrease the TFR, given that it is already very

low. However, given the levels of urbanization that can be realistically expected in

2030 and beyond, relaxing the one-child policy to a two-child policy would not have

a major effect on China’s population growth. Under this scenario, we project the

TFR to be 1.6 in 2030. This supports studies that call for alternative policies to the

one-child rule (e.g., Greenhalgh and Bongaarts 1987; Wang 2005; Zeng 2007).

Acknowledgments This study is jointly supported by Social Sciences and Humanities Research Council of Canada, Program for Chang Jiang Scholars and Innovative Research Team in Universities of

the Ministry of Education of China (IRT0855) and the National 985 Project of the Ministry of Education

and Treasury Department of China (07200701). The authors gratefully acknowledge helpful comments

from Barry Edmonston.

Appendix A: A Simplified Version of the Decomposition Equation

To derive a simplified version of decomposition equation, Eq. 1 requires three

additional assumptions: (a) Fx is constant for all x, i.e., age specific fertility rates are constant in all ages; (b) kx is constant for all x, meaning that the proportion of urban females at aged x in the total female population is constant; and (c) the sex composition of urbanites remains constant while urbanization rate (Cu) increases such that Cu = ku. Under these assumptions, Eq. 1 can be re-written as

TFRasfr ¼ FrCr þ FuCu ð3Þ

and decomposing (3),

DTFR ¼ Fu � Fr � �

DCu þ CrDFr þ CuDFu ð4Þ

where Fr and Fu denote rural TFR and urban TFR, respectively; Cr and Cu denote the proportion of rural and urban population; and again we have Cr ? Cu = 1. Table 4 shows that the difference in TFR between using Eqs. 1 and 3 is minimal (see the last

column of Table 4), suggesting that it is not unreasonable to decompose TFR, rather

than TFRasfr, in the decomposition exercise and the forecasts of TFRs (see Appendix B).

The Effect of Urbanization on China’s Fertility 431

123

Appendix B: Forecast of TFRs in Urban and Rural Areas

To forecast future national fertility, we estimated a conventional time series model

for the log-transformed rural and urban TFRs (Fr and Fu), conditional upon that the TFRs are greater than 0 (e.g., Box et al. 2008) We used data from 1950 to 2008. The

fitted models for Fu and Fr are given below (standard errors in parentheses):

ln Fu;t ¼ 0:992 0:0096ð Þ � ln Fu;t�1; R2 ¼ 0:904

ln Fr;t ¼ 0:995 0:0065ð Þ � ln Fr;t�1;R2 ¼ 0:903

Using these equations, it is straightforward to forecast TFRs in urban and rural area

for the next 20 years (see Fig. 7). Figure 7 shows that rural TFRs in next 20 years

are fairly stable at approximately 1.6, while urban TFRs are around 1.1.

Table 4 A simplified decomposition of TFRs

Year TFR (rural) TFR (urban) Urbanization rate (%) TFR (1) TFR (3) Difference

1982 3.02 1.40 21.13 2.62 2.68 0.06

1990 2.58 1.59 26.41 2.30 2.32 0.02

2001 1.60 1.08 37.66 1.39 1.40 0.01

2008 1.73 1.22 45.68 1.47 1.50 0.03

Sources: The 1982 Census, the 1990 Census; the 2001 and 2008 0.1% population surveys

0

1

2

3

4

5

6

7

8

1950 1960 1970 1980 1990 2000 2010 2020 2030

T F

R

Year

TFR(rural)

TFR(urban)

Fig. 7 Forecast of TFRs in urban and rural areas. Sources: The 1950–1977 data came from 1/1,000 fertility survey conducted in 1982; the 1978–1987 data were from the 1988 2/1000 fertility and contraceptive use survey; the 1988–1992 data came from the 1992 National Fertility Survey; the 1993–2000 data were obtained from the 2001 National Fertility and Reproductive Health Survey; and finally the 2001–2008 data came from the Annual Population Monitoring Surveys

432 Z. Guo et al.

123

References

Bongaarts, J. (2002). The end of the fertility transition in the developing world. Population Bulletin of the United Nations, 48(49), 271–286.

Bongaarts, J., & Watkins, S. C. (1996). Social interactions and contemporary fertility transitions.

Population and Development Review, 22(4), 639–682. Box, E. P., Jenkins, G. M., & Reinsel, G. C. (2008). Time series analysis: Forecasting and control. New

York: Wiley.

Bryant, J. (2007). Theories of fertility decline and evidence from development indicators. Population and Development Review, 33(1), 101–127.

Cai, Y. (2010). China’s below-replacement fertility: Government policy or socioeconomic development?

Population and Development Review, 36(3), 419–440. Das Gupta, P. (1991). Decomposition of the difference between two rates and its consistency when more

than two populations are involved. Mathematical Population Studies, 3(2), 105–125. Feeney, G., & Wang, F. (1993). Parity progression and birth intervals in China: The influence of policy in

hastening fertility decline. Population and Development Review, 19(1), 61–101. Feeney, G., & Yuan, J. (1994). Below replacement fertility in China? A close look at recent evidence.

Population Studies, 48(3), 381–394. Fu, Y., Wei, P., & Jin, R. (2009). Urbanization level in the transitional stage of China and provincial

economic growth. Paper presented at the 2009 International Conference on Management and Service Science, Wuhan, China.

Greenhalgh, S., & Bongaarts, J. (1987). Fertility policy in China: Future options. Science, 235(4793), 673–701.

Gu, B. (2007). Low fertility in China: Trends, policy, and impact. Asia Pacific Population Journal, 22(2), 73–90.

Gu, B., Wang, F., Guo, Z., & Zhang, E. (2007). China’s local and national fertility policies at the end of

the twentieth century. Population and Development Review, 33(1), 129–147. Hirschman, C. (1994). Why fertility changes. Annual Review of Sociology, 20, 203–233. Jaffe, A. J. (1942). Urbanization and fertility. American Journal of Sociology, 48(1), 48–60. Kirk, D. (1996). Demographic transition theory. Population Studies, 50(3), 361–387. Liang, Q., & Lee, C. (2006). Fertility and population policy: An overview. In D. L. Poston, C. Lee,

C. Chang, S. L. McKibben, & C. S. Walther (Eds.), Fertility, family planning, and population policy in China (pp. 8–19). New York: Routledge.

National Bureau of Statistics of China (NBS). (2009). Chinese statistics yearbook, 2009. Beijing: National Bureau of Statistics.

National Bureau of Statistics and the East-West Center. (2007). Fertility estimates for the provinces of China, 1975–2000. Beijing: National Bureau of Statistics.

Notestein, F. (1953). Economic problems of population change. Proceedings of the eighth international conference of agricultural economists (pp. 13–31). London: Oxford University Press.

Pan, J. H., & Wei, H. K. (2010). Annual report on urban development of China. Beijing: Academic Press of Social Sciences.

Peng, X. (2004). Is it time for China to change its population policy? China: An International Journal, 2(1), 135–149.

Peng, X. (2011). China’s demographic history and future challenges. Science 333 (Special section), 581–587.

Poston, D. L., & Gu, B. (1987). Socioeconomic development, family planning, and fertility in China.

Demography, 24(4), 531–551. Sharping, T. (2003). Birth control in China 1949–2000: Population policy and demographic

development. New York: Routledge Curzon. Tien, H. Y. (1984). Induced fertility transition: Impact of population planning and socio-economic change

in the People’s Republic of China. Population Studies, 38(3), 385–400. United Nations. (2010). World urbanization prospects: The 2009 revision. New York: United Nations. Wang, F. (2005). Can China afford to continue its one-child policy? Asian Pacific Issues, 77, 1–12. Wu, Z., Schimmele, C. M., & Li, S. (2009). Demographic change and economic reform. In A. Sweetman

& J. Zhang (Eds.), Economic transitions with Chinese characteristics: Social change during thirty years of reform (pp. 149–167). Kingston: McGill-Queen’s University Press.

The Effect of Urbanization on China’s Fertility 433

123

Zeng, Y. (2007). Options for fertility policy transition in China. Population and Development Review, 33(2), 215–246.

Zeng, Y., & Vaupel, J. W. (1989). The impact of urbanization and delayed childbearing on population

growth and aging in China. Population and Development Review, 15(3), 425–445. Zheng, Z. (2004). Fertility desire of married women in China. Chinese Journal of Population Studies, 5,

73–78.

434 Z. Guo et al.

123