DEPARTMENT OF ECONOMICS

ISSN 1441-5429

DISCUSSION PAPER 13/16

Fiscal Decentralisation, the Knowledge Economy and School Teachers’

Wages in Urban China*

Yi Longa, Chris Nylanda and Russell Smythb

Abstract: We examine how fiscal decentralisation and progress towards the development of a

knowledge-intensive economy has impacted on teachers’ wages in China, utilising a panel

dataset of 31 provincial administrations from 2001 to 2013. We find that fiscal

decentralisation has a negative impact on teachers’ wages and this effect is further

enhanced by a deepening of the knowledge intensity of the economy, while knowledge

economy itself has no significant impact on teachers’ wages. The findings suggest that

incentives being offered to local administrators need to be revisited if the national

government is convinced of the need to increase teacher quality in ways suited to the

knowledge economy China wishes to construct.

Keywords: fiscal decentralisation, knowledge economy, teachers, wages, human capital,

China

JEL Classification Numbers: H73, J31, J45

* We thank Roland Cheo, Zhiming Cheng and Brian Cooper for helpful suggestions on earlier versions of this

paper. We alone are responsible for the content. a Department of Management, Monash Business School, Monash University b Department of Economics, Monash Business School, Monash University

© 2016 Yi Long, Chris Nyland and Russell Smyth

All rights reserved. No part of this paper may be reproduced in any form, or stored in a retrieval system, without the prior

written permission of the author.

monash.edu/ business-economics ABN 12 377 614 012 CRICOS Provider No. 00008C

2

Introduction

China is undergoing a fundamental structural transformation from a low skilled,

surplus-labour economy to a more capital-intensive, and technology-intensive,

economy (Garnaut, 2011). To complement this shift, China’s government accepts that

it needs to increase the nation’s investment in human capital accumulation (World

Bank, 2013). This is a development that is evidenced by the marked increase in

resources now directed to education and by the increased emphasis on the need to

build a workforce that is both innovative and creative (Pang & Plucker, 2013). What

China’s national government also recognises is that socio-economic diversity

influences the extent to which regional administrations need to subsidise education

provision. This is evidenced by the fact that the national government has embraced a

regime that entails the specification of a common national minimum teacher wage

schedule and permits local authorities to decide whether, and to what extent, they will

supplement this minimum. For example, Zeng and Yu (2015) find that in 2010

regional supplements for teachers in Shanghai, Beijing, Zhejiang and Tibet were

almost four times higher than the national minimum. Regional supplements introduce

a degree of competition for teachers between provinces, although the level of regional

competition for teachers should not be overstated. While teachers are free to move

across provinces, regional administrators retain market power because teachers are

unable to transfer highly desired permanent positions across provinces.

Despite the increased value accorded to human capital development through

3

education, whether China can build a workforce suited to the utilisation of

sophisticated technology remains an open question. Heckman (2005) has charged that

China underinvested in human capital in the 1990s and that this resulted in a shortage

of workers with the cognitive capacities required to master the skills demanded by

new technologies. Consequently, there was a relatively low rate of return to the capital

that was invested in the upgraded technologies. A number of researchers have posited

that the underinvestment in education Heckman underscores was due to the system of

fiscal decentralisation that permits local governments, that tend to favour investments

that generate short term returns, to determine precisely how much will be invested in

social service provision (Fu, 2010; Luo & Chen, 2010; Wang et al., 2012).

Fiscal decentralisation is a concept that captures the extent to which national

governments permit local administrators to retain a share of their locally generated

revenue and/or decide how resources are allocated (Porcelli, 2009; Zhang & Zou,

1998). Focusing on the effect of this institutional factor on human capital formation,

this paper examines how fiscal decentralisation, and progress towards the

development of a knowledge-intensive economy, influences teachers’ wages in China.

Locating the analysis in the foregoing context, this paper makes two contributions to

teacher wage determination theory and the Chinese teacher wage determination

literature. First, it enriches our knowledge of the teacher wage determination process

by examining the role that fiscal decentralisation, and its interaction with the

knowledge economy, play as institutional determinants of teachers’ wages. Fiscal

4

decentralisation, to different degrees, is observed in many countries and is held to

influence government behaviours (OECD & Korea Institute of Public Finance, 2013).

The impact of labour market institutions, such as unions and collective bargaining, on

teacher wages has been extensively studied (Goldhaber, 2009; Hoxby, 1996).

However, although the government is a key player in the teacher wage determination

process, we really know little about the effect of the institutional incentives faced by

governments in this process. China provides the ideal setting in which to address this

issue because its fiscal reforms provide us with an opportunity to examine how

institutional incentives that governments face influence teachers’ wages.

Second, incorporating the knowledge economy into wage determination theories also

underscores the fact that the wage determination process is evolutionary. That this is

the case has been recognised by scholars who have documented how changing

attitudes to gender and race have influenced wages (Blau & Kahn, 2000; Podgursky &

Springer, 2007). But what has gone unexplored is the possibility that the changing

knowledge intensity of economies may change the context in which teachers work

and that it may be necessary to introduce wage adjustments when the labour market

increasingly rewards high skills (Acemoglu, 1998; Katz & Margo, 2014). Whether

this general trend has been reflected in a specific skill-intensive occupation, such as

teachers, is of interest and has far-reaching implications for the recruitment and

retention of high quality workers in the education sector.

To examine whether, and to what extent, fiscal decentralisation, knowledge economy

5

and their interaction affects teacher wages, we focus on China for two reasons. First,

fiscal decentralisation has been a prominent feature of China’s intergovernmental

fiscal system (Jin, et al., 2005; Zhang & Zou, 1998). While China is not unique in

experimenting with fiscal decentralisation, the fiscal reforms that China has

implemented have accentuated regional differences in the incentives faced by

governments. This provides an excellent setting in which to examine the role that the

institutional incentives faced by government play in determining the wages that

teachers receive.

A second reason for situating our study in China is that its education system has

recently been at the centre of international debates about the best way to develop a

high-performing education system. Shanghai topped the league table in the 2009 and

2012 Programme for International Student Assessment organised by the OECD,

generating considerable international attention in its approach to education. This has

also initiated fierce public debate about the average level of academic achievement

and the quality of education in China (OECD, 2011; Tucker, 2012; Zhao, 2014). There

is a need to examine how the evolution of the knowledge economy that China desires,

has effected the incentives offered to governments to attract the best people into the

education sector, not just in Shanghai, but across the country.

Teachers’ compensation includes both wages and fringe benefits. While, in the past,

many fringe benefits have not been totally transparent, recent reforms have

incorporated fringe benefits into the wage bill. For example, free housing has been

6

replaced with the housing provident fund, which is clearly listed on the wage bill. In

the past, like all civil servants, teachers did not have to contribute to social insurance,

even though they received better pensions on retirement than did employees working

in the private sector or state-owned enterprises. The pension reforms in 2014 and

2015, however, mandated that teachers now have to pay the same proportion of their

wages to the social insurance program and that the base wage should be increased to

compensate for wages contributed to social insurance. Hence, pensions are also

visible on the wage bill. These reforms suggest that differences in teachers’ wages are

the main drivers of differences in total remuneration across provinces.1

We find that fiscal decentralisation has a negative influence on teacher wages. A one

percent increase in fiscal decentralisation reduces the average annual primary school

teacher’s wages by 0.87% to 0.92% and the average annual secondary school

teacher’s wages by 0.47% to 0.52%. Finding that there is a negative link between

decentralisation and wages is important because it is commonly assumed that

devolution will enhance the extent to which wages reflect the needs of local

economies, but this may not be the case if long-term development requires a high rate

of human capital accumulation and local administrators have shorter term priorities.2

We find that progression toward a knowledge-intensive economy has no significant

1 As discussed below our measure of teachers’ remuneration covers the total wage bill, including base and

supplement wages and salaries as well as other payments both in kind and in cash. 2 One might be concerned that the negative link between fiscal decentralisation and wages is being driven by

differences in teacher quality where better qualified teachers are migrating to wealthier, higher paying, provinces

leaving less well qualified teachers in the poorer provinces. If this was the case, we would observe that wages of

teachers would be falling not because the government is paying less, but because the remaining pool are less

qualified. However, as we note above, interprovincial competition for teachers should not be overstated and, in our

modelling below, we at least partly control for differences in teacher quality through including teachers’ education

and professional rank as control variables, although neither are perfect proxies for teacher quality.

7

influence on school teacher wages. The upward pressure on teacher wages exerted by

the knowledge economy through increasing demand for skilled workers may be offset

by the rise in the supply of college graduates in China since the expansion of the

higher education sector in 1999. This explanation is in line with the observation that

the skill wage premium may not increase if the supply of educated workers outpaces

skill demand that is induced by an upgrading of technology as Goldin and Katz (2007)

find occurred in the United States through 1915-1980.

Despite finding no significant direct effect of the knowledge economy on teacher

wages, we find that progression to a knowledge-intensive economy strengthens the

negative association between fiscal decentralisation and teachers’ wages.

Conditioning on the average level of knowledge economy, a one percent increase in

fiscal decentralisation reduces the average annual primary school teacher wages by

0.94% to 1.27% and the average annual secondary school teacher wages by 0.55% to

0.92%.

While the findings generated by the study are derived from China’s experience they

have wider relevance to ongoing debates being sustained in many nations on how to

improve the quality of the teaching workforce. Giving more local and school

autonomy in wage determination have been advanced as strategies to improve the

quality of teaching (OECD, 2005; Propper & Britton, 2012). Our finding regarding

China’s experience suggests localised determination may not lead to the desired

outcome if autonomy is not balanced by appropriate incentives and accountability

8

systems.

Conceptual model

Fiscal decentralisation and teachers’ wages

That fiscal decentralisation may impact on teachers’ wages is suggested by two

strands of literature. The first argues that fiscal decentralisation tends to diminish

public services. It stems from theories of fiscal federalism which study the incentives

embedded in institutions and the types of decentralisation that best support economic

prosperity (Jin et al., 2005; Oates, 2008). China’s national government provides

strong economic and personal promotion incentives that are designed to encourage

local governments to promote economic growth and development (Chen et al., 2005;

Li & Zhou, 2005).

With autonomy to make fiscal decisions, these incentives motivate administrators to

prioritise items that contribute to short-term economic growth, which tends to mean

directing resources to build physical infrastructure that attracts investment inflow

(Cheung, 2008) and encouraging firms to take advantage of China’s abundance of

cheap labour (Gaulier et al., 2007).

In the period in which China's development strategy focussed on low value-added

production only, basic literacy was required to complement this approach. Research

for the 1990s found that providing the population with a primary school education had

a positive effect on GDP growth, but the effect of higher levels of education was

statistically insignificant (Qian & Smyth, 2006). Given this was the case and given

9

local governments were encouraged to focus on growth, it is not surprising that local

officials discounted the importance of human capital accumulation and the need to

invest in education (Gordon & Li, 2011; Li et al, 2012; Qian & Roland, 1998). This

downplayed the need to raise teachers’ wages for recruiting and retaining high skilled

staff.

In summary, this strand of thought holds that the higher the level of fiscal

decentralisation and the consequent greater fiscal autonomy, the more likely

administrators will align expenditure with fiscal and political incentives (Fu & Zhang,

2007). With a higher level of fiscal decentralisation, local government officials who

are rewarded for accelerating growth are more likely to underinvest in public service

provision and overinvest in infrastructure construction that has a significant and

immediate impact on growth.

The second strand of literature that has examined the impact of fiscal decentralisation

on local administrators suggests the priorities of the party-state have shifted over time.

In brief, it is posited that the national government has expanded its agenda and now

seeks to promote economic development and expand public service provision (Liu,

2011; Yang, 2013). Given this new mix of priorities, if incentives to encourage the

long-term development of human capital are greater than the incentives to promote

short-term growth, fiscal decentralisation will exert a positive impact on public goods

provision and, vice versa.

The overwhelming majority of empirical studies continue to find economic growth is

10

being prioritised over human capital accumulation (Fu, 2010; Jia et al., 2014; Li, 2007;

Luo & Chen, 2010; Qiao et al., 2006). It is possible that this is because most studies

use historical data even though China is continuing to change rapidly. But given the

current weight of evidence we posit the following hypothesis:

Hypothesis 1: When the level of fiscal decentralisation is higher, teachers’ wages

will be lower.

Knowledge economy and teacher wages

The possibility that progression towards a knowledge economy will have a positive

impact on teachers’ wages is suggested by Heckman’s (2005) claim that as more

sophisticated technology was introduced into Chinese industry it became necessary to

increase investment in human capital accumulation, that this was not done at a

commensurate rate, and as a consequence the return to physical capital was

diminished.

The global nature of this transformation has also been highlighted by the OECD

(1996). It has observed repeatedly that knowledge has become a primary driver of

productivity, economic growth, and competitiveness. In so doing the OECD (2008)

has also stressed that as technology becomes more sophisticated so too does the need

both for workers who can learn sophisticated skills and for teachers who have the

capacity to graduate students who have the requisite capacity to learn (see also,

Tucker, 2012).

11

Teachers’ contribution to students’ learning capacities has been evidenced by the

finding that a one standard deviation increase in teacher quality raises test scores of a

national standardised test by almost 0.1 standard deviation in reading and

mathematics in the United States (Rockoff, 2004, see also Aaronson et al., 2007;

Chetty et al., 2014a; Hanushek & Rivkin, 2010). In addition, Chetty et al. (2014b)

find students taught by high value-added teachers in primary school are more likely to

attend college and earn higher salaries. Outside the United States, research for

Australia (Leigh, 2010), Brazil (Harbison & Hanushek, 1992), the United Kingdom

(Slater et al., 2012), and Peru (Metzler & Woessmann, 2012) has confirmed that

teacher quality has a positive impact on students’ academic achievement. Thus, high

quality teachers are a prerequisite for cultivating the knowledge economy workforce.

Skill-biased technological progress increases the demand for high skilled workers and

raises their wages (Acemoglu, 1998; Katz & Margo, 2014). Thus, there is the need to

recompense teachers that will induce high quality students to become teachers and

ensure high quality teachers remain in this profession (OECD, 2008; Schleicher,

2011).

In China skill requirements have risen in line with industrial upgrading occurring at

different rates in accordance with the extent to which regions have developed a

knowledge-intensive economy (Cai & Wang, 2014; Heckman & Yi, 2012). We

suggest that given the evidenced link between the quality of teachers and wages there

will also be a positive association between wages and the extent to which regions

12

have been able to develop a knowledge intensive economy. This suggests our second

hypothesis.

Hypothesis 2: When the level of knowledge economy is higher, teacher wages

will be higher.

The moderating role of the knowledge economy on the relationship between fiscal

decentralisation and teachers’ wages

That local government fiscal independence arguably induces local administrators to

curtail teachers’ wages begs the question: is this tendency mitigated by the advance of

the knowledge economy? In the early stages of the transition to a knowledge-intensive

economy this mitigation effect is likely to occur only at the level of primary school

education given that a primary school education is sufficient for the overwhelming

majority of workers to undertake their employment. As the level of technology within

the economy increases, along with increased demand for more skilled workers and for

higher quality teachers, this effect will tend to become pronounced at higher levels of

education and local administrators will come under mounting pressure to increase the

proportion of their expenditure that they apportion to education and to the

employment of ever higher quality teachers.

The OECD (2015) and the National Centre on Education and the Economy in the

United States (Tucker, 2012) has observed that high-performing education systems

are inclined to prioritise higher teacher salaries over other forms of education

13

expenditure. Dolton and Marcenaro-Gutierrez (2011) have lent support to this claim

by evidencing that teachers’ wages, both measured in absolute terms and relative to

the average wages in a country, are positively associated with student performance

even after controlling for country fixed effects.

The development of a knowledge economy in China creates an imperative for

governments to accept the policy implications drawn from the experience of

high-performance education systems. Ramesh (2013) argues that progression towards

a knowledge-based economy is the product of a cumulative and overlapping set of

reforms which over 30 years have embedded knowledge creation in coastal regions

with the process progressively diffusing to the rest of China. Central to this process,

Ramesh (2013) adds, is education policy and more specifically the government’s

effort to raise the quality of the nation’s education workforce. If local governments

embrace this insight, we can expect the knowledge economy to moderate the negative

impact of fiscal decentralisation on teachers’ wages.

Hypothesis 3: The level of knowledge economy moderates the relationship

between the degree of fiscal decentralisation and teacher wages, such that when

the level of knowledge economy increases, the negative impact of fiscal

decentralisation on teacher wages is weakened.

Model specification, measurement of variables and dataset

The basic specification of the model can be represented by:

14

𝑊𝑖𝑡 = α𝑖 + β1FD𝑖𝑡 + β2KE𝑖𝑡 + ∑𝑋𝑖𝑡 + 𝑖𝑡

This unconditional model will be employed to test the first and second hypothesis.

The dependent variable 𝑊𝑖𝑡 is the average teacher wages in region 𝑖 in year t. The

major explanatory variables are FD𝑖𝑡, representing fiscal decentralisation, and KE𝑖𝑡

denoting knowledge economy. In addition, we also control for variables that have

been deemed important in the extant teacher wage determination literature. The set of

control variables 𝑋𝑖𝑡 include local government’s fiscal capacity, teachers’ human

capital and other demographic attributes, job attributes, changes in student enrolments,

cost of living (COL) and local unemployment rate. The intercept i may be the same

or varying across regions. The error term 𝑖𝑡 is assumed to have a zero mean, and to

be independent and identically distributed across observations.

To test whether the knowledge economy moderates the effect of fiscal decentralisation

on teachers’ wages, the interaction term of the two independent variables FD𝑖𝑡 ∗ KE𝑖𝑡

is added to the basic model. Below shows the conditional model:

𝑊𝑖𝑡 = α𝑖 + β1‘ FD𝑖𝑡 + β2

‘ KE𝑖𝑡 + β3′ FD𝑖𝑡 ∗ KE𝑖𝑡 + ∑

’𝑋𝑖𝑡 + µ𝑖𝑡

Fiscal decentralisation has been measured using various approaches

(Martinez-Vazquez & Timofeev, 2010). The most commonly used measurement in

studies for China since the mid-1990s has been the ratio of subnational government

expenditure to national government expenditure (Fu & Zhang, 2007; Li, 2007; Zhang

& Zou, 1998; Zheng, 2008). A similar measure that has been employed is the ratio of

15

subnational government revenue to national government revenue (Guo & Jia, 2010; Li,

2007; Liang, 2009; Shen & Fu, 2005). Another often utilised measurement is the

fiscal autonomy indicator or self-reliance ratio, which is derived by dividing total

subnational fiscal revenue by subnational fiscal expenditure (Chen, 2010; Gong & Lu,

2009; Xing & Deng, 2012).

Despite their widespread usage, the first two measures are problematic since a

common denominator across provinces for any given year merely captures the amount

of provinces’ public revenue or expenditure, instead of the level of fiscal

decentralisation (Chen & Gao, 2012; Niu, 2013; Zhang, 2011). Chen and Gao (2012)

and Zhang (2011) point out the choice of fiscal decentralisation measurement is

dependent on the time period covered in the study and fiscal autonomy is a better

indicator to reflect regional fiscal decentralisation variation after 1994 compared to

revenue and expenditure measurement. Thus, we use the third measure in this study.

We focus on three dimensions of the knowledge economy, based on the work of

Ramesh (2013); namely, research and development (R&D), size of the knowledge

economy workforce and knowledge creation, measured by full-time equivalent R&D

personnel per 10,000 people, population with a college or above education as a

percentage of the working age population and certified patent applications.

The dataset consists of a 13-year panel for the 31 provinces and municipalities

directly under the national government in Mainland China from 2001 to 2013. School

teachers’ average wages are obtained from the China Labour Statistical Yearbook,

16

which only has data for those working in urban areas.3 As the average wage of

teachers in primary education and secondary education are given separately, we

consider these groups independently. Table A1 in the appendix contains details of

measurement and data sources. The descriptive statistics are given in Table 1.

Table 1. Descriptive statistics for all variables

a. Wages and fiscal capacity data are adjusted for inflation by converting the figures to 2013 RMB

constant prices using the national consumer price index (CPI) as the deflator. The reason that we

did not use provincial CPI is that COL as measured by the spatial price index has captured the

regional price variation. So we only have to adjust for inflation over time.

b. P and S in the parentheses refer to primary schools and secondary schools respectively.

c. KE refers to knowledge economy

3 According to the explanatory notes in the China Labour Statistical Yearbook, “the average wage of employees

refers to the average wage level in monetary terms per employee in a certain period of time, which is calculated by

dividing the total wage bill for employees in that period by the average number of employees in that period. The

calculation of the total wage bill is based on the total remuneration payment, including wages, salaries and other

payments to employees, regardless of its source or category, both in kind and in cash”.

Variable Observations Mean SD Min Max

Wages (P) 403 31145 16469 9448 90950

Wages (S) 403 33764 16572 10869 94669

FD 403 47.96 18.67 5.07 87.23

R&D 403 15.89 19.87 0.83 114.50

Size of KE workforce 403 10.28 6.72 0.36 48.22

Knowledge creation 403 14564 32123 7 269944

Fiscal capacity 403 2803 3048 319 17501

Female (P) 403 70.45 7.53 50.21 83.14

Female (S) 403 50.04 10.86 30.74 73.82

Education (P) 403 14.42 0.72 12.06 15.89

Education (S) 403 15.65 0.27 14.94 16.41

Senior rank (P) 403 49.84 12.24 14.62 80.46

Senior rank (S) 403 53.95 11.30 16.47 81.44

Ratio (P) 403 18.56 2.66 10.97 25.43

Ratio (S) 403 16.40 2.55 9.62 22.58

Enrolment growth (P) 403 -2.28 4.49 -17.55 40.84

Enrolment growth (S) 403 0.12 5.55 -17.34 29.92

COL 403 3289 447 2430 4614

Unemployment rate 403 3.70 0.71 1.2 7.1

17

Results

The estimated direct effect of fiscal decentralisation and knowledge economy on

teachers’ wages in primary schools and secondary schools are reported in column 1 to

3 and column 4 to 6 in Table 2 respectively. As the results are similar across the two

sub-sectors in education, we discuss these two sets of results together.

Table 2. The direct effect of fiscal decentralisation and knowledge economy on teachers’ wages

a. Robust standard errors in parentheses, *** p<0.01, ** p<0.05, * p<0.1

b. The Breusch and Pagan LM test rejects the pooled OLS estimator as being unbiased and

efficient and suggests the existence of individual effects. We select the fixed effect model over the

random effect model to address the individual effects. When the unit of observation is a large

geographical unit such as state or province, we cannot treat our sample as a random sample from a

large population (Wooldridge, 2012).

c. Modified Wald test for groupwise heteroscedasticity and Wooldridge test for autocorrelation

in panel data suggest the existence of heteroscedasticity and autocorrelation in the residuals. Thus,

we clustered the standard error on the panel variable (province), producing an estimator robust to

cross-sectional heteroscedasticity and serial correlation.

Fiscal decentralisation has a negative impact on primary and secondary school

teachers’ wages, consistent with our first hypothesis. It suggests that where the level

of fiscal decentralisation is higher, teacher’s wages tends to be lower, confirming the

theoretical argument that fiscal decentralisation will underplay investment in human

VARIABLES (1) (2) (3) (4) (5) (6)

FD -274.6*** -287.6*** -270.1*** -165.4*** -175.7*** -160.1***

(47.87) (50.58) (47.08) (55.24) (52.51) (57.02)

R&D 64.21 118.9

(82.86) (99.96)

Size of KE

workforce

-174.3 -115.7

(139.5) (150.3)

Knowledge

creation

0.0144 0.0214

(0.0104) (0.0154)

Observations 403 403 403 403 403 403

Within R2 0.952 0.952 0.952 0.958 0.957 0.958

F-statistics 518.10 473.98 406.61 488.85 337.42 306.84

18

capital (Gordon & Li, 2011; Li et al., 2012; Qian & Roland, 1998).

In addition, the absolute value of the coefficients on fiscal decentralisation for

primary school teachers in all three equations are uniformly higher than those for

secondary school teachers, implying that while local governments undermine teacher

wages, they set even lower value on primary school teachers. The regression

coefficient suggests that for a one percent increase in fiscal decentralisation, the

reduction of average primary school teachers’ wages per year ranges between 270.1

RMB and 287.6 RMB and that of average secondary school teachers’ wages ranges

between 160.1 RMB and 175.7 RMB depending on the measurement of the

knowledge economy, which account for 0.87% to 0.92% of the average annual

primary school teacher’s wages and 0.47% to 0.52% of the average annual secondary

school teacher’s wages. While we do not report the results of the controls, this is

equivalent to a decrease in years of schooling of primary school and secondary school

teachers of 0.058 to 0.062 years and 0.011 to 0.013 years. In other words, the impact

of a one standard deviation increase in fiscal decentralisation on teachers’ wages is

equal to that of a 1.50 to 1.60 standard deviation decrease and a 0.76 to 0.89 standard

deviation decrease in years of schooling of primary school and secondary school

teachers, respectively.

However, the level of knowledge economy is found to have no significant influence

on primary and secondary school teacher wages, consistent across all three measures

of knowledge economy. The results reject our second hypothesis which posits a

19

positive relationship between the knowledge economy and school teacher wages.

Table 3. The effect of fiscal decentralisation on teachers’ wages conditioned on the knowledge

economy

Notes: See notes to Table 2

The interaction between fiscal decentralisation and knowledge economy has a

negative impact on both primary and secondary school teachers’ wages, which means

that the knowledge economy strengthens the negative influence of fiscal

decentralisation on teachers’ wages, inconsistent with the third hypothesis which

posits that a higher level of knowledge economy weakens the negative impact of

fiscal decentralisation on teachers’ wages. As is indicated in Figures 1 and 2, when the

level of knowledge economy increases4, the negative effect of fiscal decentralisation

4 Here we only show the graph with the size of the knowledge economy workforce as the measurement of

knowledge economy. Other graphs are available upon request.

VARIABLES (1) (2) (3) (4) (5) (6)

FD -232.0*** -234.5*** -258.3*** -139.6** -141.7*** -154.5**

(57.17) (57.82) (48.21) (58.64) (48.11) (57.84)

R&D 882.5*** 992.3***

(265.4) (249.2)

Size of KE

workforce

607.1** 426.8*

(225.0) (245.0)

Knowledge

creation

0.201** 0.192**

(0.0819) (0.0899)

FD* R&D -10.23*** -10.68***

(3.396) (3.166)

FD* Size of KE

workforce

-14.62*** -10.32*

(4.366) (5.301)

FD* Knowledge

creation

-0.233** -0.212*

(0.103) (0.112)

Observations 403 403 403 403 403 403

Within R2 0.956 0.954 0.953 0.962 0.958 0.958

F-statistics 537.22 582.34 427.51 1145.45 659.91 334.42

20

on both primary and secondary school teachers’ wages becomes larger.

At the average level of knowledge economy, as measured by the size of the

knowledge economy workforce, a one percent increase in fiscal decentralisation

reduces primary school teachers’ wages by 384.7 RMB and secondary school teachers’

wages by 247.8 RMB, which are 1.24% and 0.73% of the average annual primary and

secondary school teachers’ wages, respectively. And when the level of knowledge

economy decreases by one standard deviation, a one percent increase in fiscal

decentralisation reduces primary school teachers’ wages by 286.5 RMB and

secondary school teachers’ wages by 178.5 RMB, which are 0.92% and 0.53% of the

average annual primary and secondary school teachers’ wages, respectively. When the

level of knowledge economy increases by one standard deviation, a one percent

increase in fiscal decentralisation reduces primary school teachers’ wages by 482.9

RMB and secondary school teachers’ wages by 317.2 RMB, which are 1.55% and

0.94% of the average annual primary and secondary school teachers’ wages,

respectively.

21

Figure 1. Conditional marginal effects of fiscal decentralisation on primary school teachers’ wages with

95% confidence interval

Figure 2. Conditional marginal effects of fiscal decentralisation on secondary school teachers’ wages

with 95% confidence interval

We calculated the total effect of fiscal decentralisation which sums the coefficients on

fiscal decentralisation and the interaction between fiscal decentralisation and the

knowledge economy evaluated at the average level, expressed as β1‘ +β3

‘ 𝐾𝐸̅̅̅̅ 𝑖𝑡. The

22

results are shown in Table 4. The total effect of fiscal decentralisation on primary

school teachers’ wages ranges between 292.29 RMB and 394.6 RMB, which are 0.94%

to 1.27% of the average annual primary school teacher wages, depending on the

specification of knowledge economy. The negative effect of fiscal decentralisation on

teachers’ wages is smaller in secondary schools, which are in the range 185.4-309.3

RMB, accounting for 0.55% to 0.92% of the average annual secondary school teacher

wages.

Table 4. Total effect of fiscal decentralisation on teachers’ wages

The national government introduced wage reform for teachers in 2009 and required

subnational government to pay local supplements to teachers. Does the effect of FD

and KE on wages change before and after the 2009 teacher wage reform? We created

a reform dummy variable, with the value zero for the period 2001 to 2009 and 1 for

the period 2010 to 2013, since Pang et al. (2010) observed that this reform was not in

effect in most provinces until 2010. First, we added the reform dummy and its

interaction with each independent variable in the unconditional model to ascertain

whether the direct effect of FD and KE changes and then, second, to the conditional

Measurement of KE dy/dx SE z P>z

Primary

school

teacher

R&D -394.6 74.1 -5.32 0.000

Size of KE workforce -384.7 50.9 -7.56 0.000

Knowledge creation -292.3 45.0 -6.50 0.000

Secondary

school

teachers

R&D -309.3 68.2 -4.53 0.000

Size of KE workforce -247.8 69.9 -3.55 0.000

Knowledge creation -185.4 56.5 -3.28 0.001

23

model to check whether the interaction effect of FD and KE changes.

When we interacted the reform dummy with FD in the unconditional model we found

that the interaction term was insignificant, suggesting that our results are not affected

by the reform. Moreover, the interaction of the reform and KE was insignificant as

well in five out of six regressions. Only in the secondary school teacher wages

regression, when KE is measured by knowledge creation, did the interaction term

become significant with a negative sign. When we interacted the reform dummy with

the interaction of FD and KE in the conditional model, the interaction term was

insignificant in five out of six regressions; the exception being the secondary school

teacher wages regression when KE is measured by R&D.

An alternative explanation for our results may be that FD has a negative impact on

local government's expenditure on both infrastructure and public service provision,

resulting in lower teacher wages in higher FD regions. To test this explanation, we

examined whether FD influenced expenditure on infrastructure, replicating the

frequently cited study by Fu and Zhang (2007). The dependent variable is the

percentage of expenditure on capital construction in total government expenditure of a

province, which is regressed on FD and a series of controls. The results were mixed

depending on the measure of FD. Employing the measure of FD used by Fu and

Zhang (2007), which is the ratio of subnational government expenditure per capita to

national government expenditure per capita, we found that FD is positive, consistent

with our theoretical argument that local governments with higher FD divert fiscal

24

resources away from public service provision and into infrastructure, and thus lower

teachers’ wages. However, as discussed above, this measure of FD has been criticised

on the basis that the denominator is the same for all provinces in a given year (see eg.

Chen & Gao, 2012; Niu, 2013; Zhang, 2011). When we use the ratio of local fiscal

revenue to local fiscal expenditure to measure FD, the coefficient on FD is

insignificant. Hence, the effect of FD on infrastructure is either positive or

insignificant, casting doubt on this alternative explanation for our results.

Discussion of findings

The study has found that fiscal decentralisation has a negative impact on teachers’

wages. This finding resonates with existing theoretical arguments and empirical

evidence which contends that in China fiscal decentralisation commonly undermines

investment in social service provision (Fu, 2010; Luo & Chen, 2010; Wang et al.,

2012). The finding is a source of concern given that Heckman (2005) has shown

China has a history of underinvestment in the development and accumulation of

human capital. It is also disconcerting given that the national government is seeking to

encourage local administrators to expand their foci to include both short term

economic growth and long term development and social stability by increasing the

provision of social services and the resources required to generate the accumulation of

high quality human resources.

In choosing to examine the association between fiscal decentralisation and teachers’

wages and by showing that this form of decentralisation may impact wages this paper

25

has expanded the variables that are normally considered when theorising the teacher

wage determination process. This is a contribution that has pertinence both for China

and for the teacher wage determination literature more broadly as fiscal

decentralisation, to varying degrees, is the norm in many nations. That the study has

shown that fiscal decentralisation has a negative impact on teachers’ wages in China

is also a significant finding. This revelation extends the extant literature on fiscal

decentralisation and government expenditure and shows that this contingency can

have a decided impact on teachers’ wages. In the study the focus has been directed to

the wages of teachers, but it is highly plausible that a similar relationship may exist in

relation to other public service occupations across China.

That the study finds that there is no association between teachers’ wages and the

intensity of the knowledge economy across China’s regions is somewhat unexpected.

The public policy literature argues consistently that increasing the knowledge

intensity of a nation or a region’s economy will normally require an increase in

investment in education (OECD, 2008; Schleicher, 2006; World Bank, 2003). It also

argues that high wages can make an important contribution to the quality of teaching.

This is not least by increasing the quality of the high school graduates who become

teachers and by inducing good teachers to remain in the field (OECD, 2005). But

whether the association between the knowledge intensity of economies and teachers’

wages exists has not previously been documented.

Ramesh (2013) argues that China has facilitated the development of a knowledge

26

intensive economy by establishing high-technology parks, encouraging research, and

expanding the provision of education services. But he suggests that Confucian based

methods of teaching and learning prevent innovation. Consequently, China’s

government needs to find ways to produce teachers who can embrace forms of

pedagogy that are better suited to the generation of graduates who are critical,

innovative and able to learn the skills they will need to effectively participate in a

knowledge intensive economy. This need is likely to be assisted by raising teachers’

wages as the economy becomes more knowledge intensive. We, however, do not find

a positive association between the knowledge intensity of economies and teachers’

wages in China.

A possible and also reasonable explanation for this result is that an increasing demand

for high skilled labour is balanced by a rising supply of such workers. According to

Goldin and Katz (2007), the skill wage premium will not increase if the supply of

educated workers outpaces skill demand due to the upgraded technology. They find

that the wage return to education initially peaked in the United States in 1915 and did

not return to this peak until the 1980s. At the latter date the increasing demand for

skilled labour began to outpace the increased supply of educated workers and as this

situation unfolded the wage return to knowledge began to increase. In China, the

situation is similar to what occurred in the United States between 1915 and the 1980s.

Li et al. (2011) argue that China intentionally prioritised the development of tertiary

education over primary and secondary education and this created significant problems

27

of absorption and unemployment for tertiary education graduates in the labour market.

As is shown in Figure 3, the number of tertiary education graduates in China

increased dramatically from 2001 to 2013. This pushed up the supply of highly

educated workers and put downward pressure on the wages of these workers. In this

environment an increase in demand for highly skilled teachers was likely to be met by

the increasing supply of graduates and consequently the growth of the knowledge

economy was not associated with increases in teachers’ wages.

Figure 3. The number of tertiary education graduates in China from 1995 to 2013 (unit, 10,000 people)

Our finding that the development of a knowledge economy in China’s regions has not

mitigated the negative effect of fiscal decentralisation is both surprising and

disconcerting. This finding suggests that while administrators who have experienced a

deepening in the knowledge intensity of their regional economy may have become

prone to increase investment in human capital accumulation they may have also

0

100

200

300

400

500

600

700

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

28

gained kudos for realising a goal favoured by the national government. If this is the

case they may also have gained an increase in their capacity to exercise autonomy and

succumb to pressures to favour short-term growth.

It is also possible that our finding may reflect the fact that determining a wage that

will balance the need to contain education expenditures while concomitantly building

a teaching labour force that can graduate students with the learning capacities of a

future society is bound to be difficult. So doing requires that administrators accurately

forecast the attributes required of teachers in 15-20 years and pay current teachers a

wage that will induce today’s high school graduates to become educators who can

garner these skills. Setting teachers’ wages at a level that is commensurate with

economies that are becoming more knowledge intensive requires, moreover, that local

governments not succumb to the temptation to concentrate education expenditure in

areas that are easily measured or are highly prestigious. That local governments are

likely to fall prey to this last possibility has been underscored by Hanushek &

Woessmann (2015) who have documented the marked tendency for administrators to

invest in measures such as years of school attainment, class size reduction, or number

of tertiary institutions established rather than on raising teacher quality. The latter

possibility is disconcerting given that recent studies suggest education quality matters

more than quantity and that the contribution of tertiary education to economic growth

disappears when the quality of schooling is considered in growth models (Hanushek

& Woessmann, 2015; OECD, 2015).

29

Conclusion

This paper has provided evidence that fiscal decentralisation does have a significant

effect on urban school teachers’ wages in China. It finds that fiscal decentralisation

has a negative impact on teachers’ wages. But unexpectedly, our research does not

find that the growth of a knowledge economy in China is an important teacher wage

determinant.

Although there seems no direct effect of knowledge economy on teacher wages, the

negative impact of fiscal decentralisation has been strengthened by knowledge

economy. We have proposed possible explanations for this finding in the discussion

section. Whether this is the case awaits testing.

These findings suggest that local governments in China have not sufficiently modified

their inclination to concentrate on short-term growth. Given this is the case we

conclude that that if the national government is convinced of the need to increase

teacher quality it will need to revisit the incentives being offered to local

administrators to invest in the development of human capital and embrace wage

policies that are suited to the knowledge economy China wishes to construct.

30

References:

Aaronson, D., Barrow, L., & Sander, W. (2007). Teachers and student achievement in

the Chicago public high schoos. Journal of Labor Economics, 25(1), 95-135.

Acemoglu, D. (1998). Why do new technologies complement skills? Directed

technical change and wage inequality. Quarterly Journal of Economics, 113,

1055-1090.

Blau, F. D., & Kahn, L. M. (2000). Gender differences in pay. Journal of Economic

Perspectives, 14(4), 75-99.

Brandt, L., & Holz, Carsten A. (2006). Spatial price differences in China: Estimates

and implications. Economic Development and Cultural Change, 55(1), 43-86.

Cai, F., & Wang, M. (2014). Accumulate human capital for China's sustainable growth.

In F. Cai (Ed.), Chinese Research Perspectives on Population and Labor,

Volume 1. Leiden: Brill.

Chen, S. (2010). Fenshuizhi gaige, difang caizheng zizhuquan yu gonggongpin gongji

[Tax-share reform, local fiscal autonomy, and public goods provision]. China

Economic Quarterly, 9(4), 1427-1446.

Chen, S., & Gao, L. (2012). Yangdiguanxi: Caizhengfenquan duliang ji zuoyongjizhi

zaipinggu [Central-local relations: Evaluation on the measurement of fiscal

decentralisation and its impact]. Management World, (6), 43-59.

Chen, Y., Li, H., & Zhou, L.-A. (2005). Relative performance evaluation and the

turnover of provincial leaders in China. Economics Letters, 88(3), 421-425.

Chetty, R., Friedman, J. N., & Rockoff, J. E. (2014a). Measuring the impacts of

teachers I: Evaluating bias in teacher value-added estimates. American

Economic Review, 104(9), 2593-2632.

Chetty, R., Friedman, J. N., & Rockoff, J. E. (2014b). Measuring the impacts of

teachers II: Teacher value-added and the student outcomes in adulthood.

American Economic Review, 104(9), 2633-2679.

Cheung, S. N. S. (2008). The economic system of China. Paper presented at the

Conference on China's Economic Transformation, Chicago.

Dolton, P., & Marcenaro-Gutierrez, O. D. (2011). If you pay peanuts do you get

monkeys? A cross-country analysis of teacher pay and pupil performance.

Economic Policy, 26(65), 5-55.

Fu, Y. (2010). Caizhengfenquan, zhengfuzhili, yu feijingjixing gonggongwupin gongji

[Fiscal decentralisation, governance, and non-economic public goods

provision]. Economic Research Journal, (8), 4-15.

Fu, Y., & Zhang, Y. (2007). Zhongguoshi fenquan yu caizhengzhichujiegou pianxiang:

Wei zengzhang er jingzheng de daijia [Chinese style decentralisation and

distorted fiscal expenditure structure: The cost of competing for economic

growth]. Management World, (3), 4-12.

Garnaut, R. (2011). Australian opportunities through the Chinese structural

transformation. Australian Economic Review, 44(4), 437-445.

31

Gaulier, G., Lemoine, F., & Ünal-Kesenci, D. (2007). China’s integration in east Asia:

Production sharing, FDI & high-tech trade. Economic Change and

Restructuring, 40(1-2), 27-63.

Goldhaber, D. (2009). The politics of teacher pay reform. In M. G. Springer (Ed.),

Performance incentives: Their growing impact on American K-12 education

(pp. 25-42). Washington, DC: Brookings Institution Press.

Gong, F., & Lu, H. (2009). Gonggongzhichujiegou, pianhaopipei yu caizhengfenquan

[Public expenditure components, preference-matching and fiscal

decentralisation]. Management World, (1), 10-21.

Goldin, C., & Katz, L. F. (2007) The race between education and technology: The

evolution of U.S. educational wage differentials, 1890 to 2005. NBER

Working Paper No. 12984. National Bureau of Economic Research.

Cambridge, MA.

Gordon, R. H., & Li, W. (2011). Provincial and local governments in China: Fiscal

institutions and government behavior. NBER Working Paper No.16694.

National Bureau of Economic Research. Cambridge, MA.

Guo, Q., & Jia, J. (2010). Caizhengfenquan, zhengfu zuzhijiegou yu difangzhengfu

zhichuguimo [Fiscal decentralisation, government structure and local

government's expenditure size]. Economic Research Journal, 11, 59-72.

Hanushek, E. A., & Rivkin, S. G. (2010). Generalizations about using value-added

measures of teacher quality. The American Economic Review, 100(2), 267-271.

Hanushek, E. A., & Woessmann, L. (2015). The knowledge capital of nations:

Education and the economics of growth. Cambridge, MA: The MIT Press.

Harbison, R. W., & Hanushek, E. A. (1992). Educational performance of the poor:

Lessons from rural northeast Brazil. New York: Oxford University Press.

Heckman, J. J. (2005). China's human capital investment. China Economic Review,

16(1), 50-70.

Heckman, J. J., & Yi, J. (2012). Human capital, economic growth, and inequality in

China. NBER Working Paper No.18100. National Bureau of Economic

Research. Cambridge, MA.

Hoxby, C. M. (1996). How teachers' unions affect education production. The

Quarterly Journal of Economics, 111(3), 671-718.

Jia, J., Guo, Q., & Zhang, J. (2014). Fiscal decentralisation and local expenditure

policy in China. China Economic Review, 28, 107-122.

Jin, H., Qian, Y., & Weingast, B. R. (2005). Regional decentralisation and fiscal

incentives: Federalism, Chinese style. Journal of Public Economics, 89(9–10),

1719-1742.

Katz, L. F., & Margo, R. A. (2014). Technical change and the relative demand for

skilled labor: The United States in historical perspective. In L. P. Boustan & C.

M. Frydman, Robert A. (Eds.), Human capital in history: The American

record (pp. 15-57). Chicago: University of Chicago Press.

Leigh, A. (2010). Estimating teacher effectiveness from two-year changes in students’

test scores. Economics of Education Review, 29(3), 480-488.

32

Li, H., & Zhou, L.-A. (2005). Political turnover and economic performance: The

incentive role of personnel control in China. Journal of Public Economics,

89(9–10), 1743-1762.

Li, J., Zhang, Q., & Wang, X. (2012). Caizhengfenquan yu touzi pianhao de

difangzhengfu xingwei: Jiyu shengji mianbanshuju de shizheng yanjiu [Fiscal

decentralisation and local government behavior of investment bias: an

empirical study based on provincial panel data]. Industrial Economics

Research, 60(5), 72-79.

Li, W. (2007). Caizhengfenquan yu difangzhengfu zhichujiegou pianxiang: Jiyu

zhongguo shengji mianbanshuju de yanjiu [Fiscal decentralisation and fiscal

expenditure composition of subnational governments: based on China's

provincial panel data]. Journal of Shanghai University of Finance and

Economics, 9(5), 75-82.

Li, Y. A., Whalley, J., Zhang, S., & Zhao, X. (2011). The higher educational

transformation of China and its global implications. The World Economy,

34(4), 516-545.

Liang, R. (2009). Caizhengfenquanxia de jinshengjili, bumenliyi yu tudiweifa

[Promotion incentives, departmental interests and land law breaking under

fiscal decentralisation]. China Economic Quarterly, 9(1), 283-306.

Liu, T.-w. (2011). The politics of social spending in China: The role of career

incentives. (PhD thesis), University of California, San Diego, the United

States.

Luo, W., & Chen, S. (2010). Fiscal decentralisation and public education provision in

China. Canadian Social Science, 6(4), 28-41.

Martinez-Vazquez, J., & Timofeev, A. (2010). Decentralisation measures revisited.

Public Finance and Management, 10(1), 13-47.

Metzler, J., & Woessmann, L. (2012). The impact of teacher subject knowledge on

student achievement: Evidence from within-teacher within-student variation.

Journal of Development Economics, 99(2), 486-496.

Niu, M. (2013). Fiscal decentralisation in China revisited Australian Journal of Public

Administration, 72(3), 251-263.

Oates, W. E. (2008). On the evolution of fiscal federalism: Theory and institutions.

National Tax Journal, 61(2), 313-334.

OECD. (1996). The knowledge-based economy. Retrieved from

http://www.oecd.org/sti/sci-tech/1913021.pdf

OECD. (2005). Teachers matter: Attracting, developing and retaining effective

teachers. Paris: OECD publishing.

OECD. (2008). Innovating to learn, learning to innovate. Paris: OECD Publishing.

OECD. (2011). Strong performers and successful reformers in education: Lessons

from PISA for the United States. Paris: OECD Publishing.

OECD. (2015). Universal basic skills: What countries stand to gain. Paris: OECD

Publishing.

OECD, & Korea Institute of Public Finance. (2013). Measuring fiscal

33

decentralisation: Concepts and policies. Paris: OECD Publishing.

Pang, L., Han, X., Xie, Y., Li, L., & Xia, J. (2010). Wanshan jizhi, luoshi yiwujiaoyu

jiaoshi jixiaogongzi zhengce [Improving mechanism to implement the policy

of performance-related pay of compulsory education teachers]. Educational

Research, 363(4), 40-44.

Pang, W., & Plucker, J. (2013). Recent transformations in China's economic, social,

and education policies for promoting innovation and creativity. The Journal of

Creative Behavior, 46(4), 247-273.

Podgursky, M., & Springer, M. G. (2007). Teacher performance pay: A review.

Journal of Policy Analysis and Management, 26(4), 909-949.

Porcelli, F. (2009). Fiscal decentralisation and efficiency of government: A brief

literature review. Department of Economics, University of Warwick, the

United Kingdom. Retrieved from

http://www2.warwick.ac.uk/fac/soc/economics/staff/fporcelli/dec_efficiency_

gov.pdf

Propper, C., & Britton, J. (2012). Does wage regulation harm kids? Evidence from

English Schools. The Centre for Market and Public Organisation, University

of Bristol, the United Kingdom. Retrieved from

www.bristol.ac.uk/cmpo/publications/papers/2012/wp293.pdf

Qian, X., & Smyth, R. (2006). Growth accounting for the Chinese provinces 1990–

2000: Incorporating human capital accumulation. Journal of Chinese

Economic and Business Studies, 4(1), 21-37.

Qian, Y., & Roland, G. (1998). Federalism and the soft budget constraint. American

Economic Review, 88(5), 1143-1162.

Qiao, B., Fan, J., & Feng, X. (2006). Fiscal decentralisation and compulsory primary

education: The case of China. Social Sciences in China: A Quarterly Journal,

27(2), 62-76.

Ramesh, S. (2013). China's transition to a knowledge economy. Journal of the

Knowledge Economy, 4(4), 473-491.

Rockoff, J. E. (2004). The impact of individual teachers on student achievement:

Evidence from panel data. The American Economic Review, 94(2), 247-252.

Schleicher, A. (2006). The economics of knowledge: Why education is key for

Europe's success. Lisbon Council Policy Brief, 1(1). Retrieved from

http://www.oecd.org/education/skills-beyond-school/36278531.pdf

Schleicher, A. (2011). Building a high-quality teaching profession: Lessons from

around the world. Paris: OECD publishing.

Shen, K., & Fu, W. (2005). Zhongguo de caizhengfenquan zhidu yu diqu

jingjizengzhang [The relationship between China's decentralised system in

finance and the regional economic growth]. Management World, (1), 31-39.

Slater, H., Davies, N. M., & Burgess, S. (2012). Do teachers matter? Measuring the

variation in teacher effectiveness in England. Oxford Bulletin of Economics

and Statistics, 74(5), 629-645.

Tucker, M. S. (2012). Researching other countries' education systems: Why it's

34

indispensable but tricky, how we did it, why this time it's different. In M. S.

Tucker (Ed.), Surpassing Shanghai: An agenda for American education built

on the world's leading systems (pp. 1-18). Cambridge: Harvard Education

Press.

Wang, W., Zheng, X., & Zhao, Z. (2012). Fiscal reform and public education spending:

A quasi-natural experiment of fiscal decentralisation in China. Publius: The

Journal of Federalism, 42(2), 334-356.

Wooldridge, J. M. (2012). Introductory econometrics: A modern approach (5th ed.).

New York: South-Western College Pub.

World Bank. (2003) Lifelong learning in the global knowledge economy: Challenges

for developing countries. Washington D.C.: The World Bank.

World Bank. (2013). China 2030: Building a modern, harmonious, and creative

society. Washington, DC: World Bank.

http://documents.worldbank.org/curated/en/2013/03/17494829/china-2030-bui

lding-modern-harmonious-creative-society

Xing, Z., & Deng, C. (2012). Caizhengfenquan yu nongcun yiwujiaoyu yanjiu: Jiyu

caizhengzizhudu shijiao [Fiscal decentralisation and rural compulsory

education: Based on self-financing rate]. Economic Issues in China, 273(4),

62-68.

Yang, L. (2013). Zhongguo ganbu guanlitizhi jianshaole difangzhengfu jiaoyuzhichu

ma? Laizi shengji guanyuan de zhengju [Does China's cadres' management

system lead to less educational spending? Evidence from provincial data].

Journal of Public Management, (2), 41-51.

Zeng, X., & Yu, X. (2015). Jiaoshi lanpishu: Zhongguo zhongxiaoxue jiaoshi fazhan

baogao (2014) [Blue Book of Teacher: Annual Report on the Teachers in

China (2014)]. Beijing: Social Sciences Academic Press.

Zhang, G. (2011). Celiang zhongguo de caizhengfenquan [Measure fiscal

decentralisation in China]. Comparative Economic and Social Systems, 158(6),

48-60.

Zhang, T., & Zou, H.-f. (1998). Fiscal decentralisation, public spending, and

economic growth in China. Journal of Public Economics, 67(2), 221-240.

Zhao, Y. (2014). Who's afraid of the big bad dragon? Why China has the best (and

worst) education system in the world. CA: Jossey-Bass.

Zheng, L. (2008). Caizhengfenquan, zhengfujingzheng yu gonggongzhichujiegou:

Zhengfu jiaoyuzhichubizhong de yingxiangyinsu fenxi [Fiscal decentralisation,

local government competition and public expenditure structure: An analysis on

the factors affecting the proportion of education expenditure], Economic

Science, (1), 28-40.

35

Appendix

Table A1. Measurement and data sources of variables

a. China Innovation Index calculated by National Bureau of Statistics of China uses 10,000

persons as the denominator- see

http://www.stats.gov.cn/english/PressRelease/201402/t20140220_513946.html for more

information.

b. Values are missing in this source, such as Shanghai 2001, Tibet 2001, 2003, 2006, 2007 and

2008. The Shanghai 2001 data was retrieved from Shanghai Statistical Yearbook 2002. And Tibet

2001 and 2003 data were retrieved from Comprehensive Statistical Data and Materials on 55

Years of New China. Tibet 2006, 2007 and 2008 data was found in the Tibet Autonomous Region

Government Report 2007, 2008 and 2009.

c. Brandt and Holz (2006) computed the index up to 2004. We updated the index to 2013

following their method.

Variable Measurement ata sources

Wages (P) Primary school staff average wages per year (RMB) China Labour Statistic Yearbook

Wages (S) Secondary school staff average wages per year (RMB) China Labour Statistic Yearbook

FD Provincial revenue/provincial expenditure (%) Finance Yearbook of China

R&D R&D personnel in full-time equivalent per 10,000 persons a China Statistical Yearbook of Science and

Technology

Size of KE workforce Population with college or above education as a percentage

of working age population (15 to 64)

China Population and Employment

Statistics Yearbook/ China Population

Statistics Yearbook

Knowledge creation Certified patent applications per year China Statistical Yearbook of Science and

Technology

Fiscal capacity Fiscal revenue per capita Finance Yearbook of China

Female (P) Proportion of female teachers in primary schools (%) Educational Statistic Yearbook

Female (S) Proportion of female teachers in secondary schools (%) Educational Statistic Yearbook

Education (P) Average years of schooling of primary school teachers Educational Statistic Yearbook

Education (S) Average years of schooling of secondary school teachers Educational Statistic Yearbook

Senior rank (P) Proportion of teachers with senior professional rank in

primary schools (%)

Educational Statistic Yearbook

Senior rank (S) Proportion of teachers with senior professional rank in

secondary schools (%)

Educational Statistic Yearbook

Ratio (P) Student-teacher ratio in primary schools Educational Statistic Yearbook

Ratio (S) Student-teacher ratio in secondary schools Educational Statistic Yearbook

Enrolment growth (P) Enrolment change rate in primary schools (%) Educational Statistic Yearbook

Enrolment growth (S) Enrolment change rate in secondary schools (%) Educational Statistic Yearbook

COL Spatial price index Brandt and Holz (2006) c

Unemployment rate Urban registered unemployment rate (%) Website of National Bureau of Statistics