# Why China is the FinTech capital of the world?

Why China is the FinTech capital of the world? Probably because no other country has poured more capital into FinTech, but a risk arises!

What is that risk?

#### China’s aging population!

It is shown that without reforms, the capital stock, labor supply and aggregate output will experience dramatic declines in the coming years of this century. The impact on return to capital and wage rate differs due to the changes in *K/L *ratio. The postponement of retirement age proves effective in delaying the rise of old-age dependency ratio, while birth control policy reform produces better results in coping with the aging population in the long run.

Acknowledgement

I would like to thank Professor Kelly Ragan for her insightful lectures on dynamic macroeconomic analysis and Teaching Assistant Matthias H¨ansel for his seminars on MATLAB skills. I would also like to thank the United Nations’ Department of Economic and Social Affairs and the World Bank for providing the data sets which make this mini-thesis possible.

1

Contents

List of Figures 3 List of Tables 3 1 Introduction 4 2 Literature Review 6

3 Model 8 3.1 Firm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3.2 Demographics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 3.3 Government . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.4 Households . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 3.5 Characteristics in equilibrium . . . . . . . . . . . . . . . . . . . . . . 10

4 MATLAB Implementation 11

5 Calibration 13 5.1 Demographics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 5.2 Preference and technology . . . . . . . . . . . . . . . . . . . . . . . . 15 5.3 Government . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 5.4 Policy reforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

6 Results 18 6.1 Time paths of individual wealth and consumption . . . . . . . . . . . 18 6.2 Properties of steady states . . . . . . . . . . . . . . . . . . . . . . . . 19

7 Conclusion 23 8 References 24 A Appendix 26

2

List of Figures

1 UN demographic projections in China . . . . . . . . . . . . . . . . . . 4 2 Population projection during Year 2020-2100 . . . . . . . . . . . . . . 15 3 Population projection with Three-child policy during Year 2020-2100 16 4 Time paths of individual capital and consumption . . . . . . . . . . . 18

List of Tables

1 Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2 Survival Probability . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 3 Reform evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3

1 Introduction

Since the implementation of the one-child policy in 1982, China’s population growth rate has experienced a decreasing trend. During the same period, the life expectancy of Chinese people has increased from 67.63 years in 1982 to 76.91 years in 2019^{1}. These two factors have resulted in the rapid aging demographics in China.

According to the United Nations’ population projection (DESA, 2019), the *old age dependency ratio *(population 65 and older divided by population 15-64) in China will increase from 12% in 2010 to 58% in 2100^{2}. The growing elderly population has posed significant challenges to the government’s ability to deliver social welfare policy as well as the long-term development of the Chinese economy.

Panel A: Total population and growth rate Panel B: Old-age dependency ratio

Figure 1: UN demographic projections in China

Source: United Nations, Department of Economic and Social Affairs, Population Division (2019). World Population Prospects 2019, Online Edition. Rev. 1.

Compared with the aging problem in other countries, there are some special characteristics in China’s problem. Firstly, the decline in population growth rate of China is not a natural process which is often the case in other countries. Its steady decrease is accompanied by the implementation of the one-child policy. By relaxing the

^{1 Data} are taken from the World Bank. See: https://data.worldbank.org/indicator/SP

.DYN.LE00.IN?end=2019&locations=CN&start=1982

^{2}The official retirement age in China currently is 60 for men, 55 for female civil servants and 50 for female workers. Suppose both men and women retire at 60, the old-age dependency ratio could be as high as 78% in 2100.

4

strict one-child policy, the long-run birth rate in China could rebound and ease the aging problem. Secondly, the official retirement age in China (women in particular) is smaller than other aging countries^{3 }and this leaves room for the Chinese government to delay retirement age in order to lower the old-age dependency ratio.

In 2016, Chinese policy makers abolished the one-child policy and introduced the two-child policy which allows two children for each couple. Besides, there are also calls for delaying retirement age and abolishing birth control policy in recent years. This thesis aims to explore the implications of the aging population in China and evaluate the potential policy measures which focus on birth control and retirement age.

The remainder of this thesis is organized as follows. Section 2 provides the literature review on aging demographics and Section 3 presents the model used in this thesis. Section 4 and 5 introduce the implementation of this model in MATLAB and its calibration. The quantitative findings are presented in Section 6 and Section 7 concludes this thesis. Appendix A gives detailed derivation of the Euler Equation and the second-order difference equation used in the backward shooting algorithm.

^{3}In 2018, OECD average retirement age is 64.2 for men and 63.5 for women(OECD, 2019). 5

2 Literature Review

The process of aging is not exclusive to China, many countries in the world have already experienced it for many years. As pointed out by Conesa and Kehoe (2018), the rapid increase of old-age dependency ratios is not only observed in developed countries, but some emerging countries are also projected to encounter this process. In this context, a large amount of literature has focused on the macroeconomic implications of aging and most studies seem to agree that there exists a negative association between aging population and economic growth.

Bernheim et al. (2001) use US data to demonstrate that there is significantly neg ative relationship between consumption and retirement, and they infer the increase of retirees would pose a negative impact on consumption and economic growth. Lee and Mason (2007) reveal that the increase of old-age dependency would raise the tax burden of the declining working-age group, which tends to have higher consumption than the old-age group. This would reduce the overall consumption of the economy and hinder its growth. Apart from consumption, the savings pattern would also be negatively impacted by the aging population. Davies and Reed (2006) show that overall saving tends to decline due to the increase of the old-age population since saving becomes the main source of spending for pensioners. Besides, there are also other routes of influencing economic growth by an aging population, such as public expenditures (Lee et al., 2011) and human capital (Bell and Rutherford, 2013).

Most of the literature focuses on research in developed countries (Teixeira et al., 2017) and as the largest developing country in the world, China has not received enough attention yet. Existing literature differs in the economic impact of the aging population in China. Modigliani and Cao (2004) argue that population aging has a negative impact on saving rates and economic growth in China. However, a recent study by Hsu et al. (2018) finds that the aging population in China would promote more savings and accelerate physical and human capital accumulation.

To deal with an aging society, many scholars have recommended two approaches: (i) the official retirement age must be delayed (Heijdra and Romp, 2009) and (ii) the fertility rate must be improved above the replacement level fertility^{4}. The Institute

^{replacement} fertility is the total fertility rate at which women give birth to enough babies to sustain population levels.

6

of Medicine (US) Committee on the Long-Run Macroeconomic Effects of the Aging U.S. Population (Council, 2013) points out that retirement age delay is the short term solution to aging problem, and fertility rate improvement would be the long term remedy after conducting a large-scale study on the implications of demographic shift in the US.

Lastly, to evaluate the impact of policy in dealing with an aging population, many researchers employ models with overlapping generations. These models range from stylised 2-3 period representations of the life cycle that build on Samuelson (1958) and Diamond (1965) work, to detailed multi-period models that build on Auerbach and Kotlikoff (1987) 65-period model.

7

3 Model

The model of this thesis is based on Problem Set 3. I choose an overlapping generation (OLG) model to evaluate the impact of the potential policy reforms the government could implement.

3.1 Firm

There is a representative firm that produces a single good using a Cobb-Douglas production function:

*Y** _{t }*=

*A*

_{t}*K*

^{θ}

_{t }*L*

^{1}

^{−θ}

*t *

where *θ *is the output share of capital, *K** _{t }*and

*L*

*are the capital and labor input at time t, and*

_{t }*A*

*is the total factor productivity at time t.*

_{t}Capital depreciation rate is *δ ∈ *(0*, *1). The representative firm maximizes its profits such that the rental rate of capital, *r** _{t}*, and the wage rate

*w*

*, are given by:*

_{t}*r** _{t }*=

*θA*

_{t}

^{ }*K*_{t}* **L*_{t}* *

_{θ−}_{1}

*− δ *(1)

_{θ}* *

(2)

3.2 Demographics

*w** _{t }*= (1

*− θ*)

*A*

_{t}

^{ }*K*_{t}* **L*_{t}* *

This model treats one period as a five-year interval and individuals enter the economy at the age of 20. After entering the economy, they will face mortality risk and work up until the mandatory retirement age of 60 ^{5}, after which individuals can live up to 100 years old. Therefore, the maximum periods *T *individuals live up to is 16 periods _{5}) in this model, during which the working life *R *lasts 8 periods ( ^{60}^{−}^{20}

(^{100}^{−}^{20}

_{5}) and

The rest periods are their retirement life. The population of each generation is given

by:

*N*_{t }_{=}X^{T}* *

*i*=1

^{ }Y *i−*1

*j*=0

*ψ*_{j}* *

!

*N** _{i,t }*(3)

where *ψ** _{j}*is the conditional probability of surviving to period

*j*+ 1, and

*N*

*is the number of individuals at the age of*

_{i,t }*i*who are born in time

*t*.

^{5}Here I ignore the differences in official retirement age between men and women in China and assume everyone retires at the age of 60 in the benchmark scenario.

8

The size of the population evolves over time exogenously at the rate *g** _{t}*, so

*N*

_{t}_{+1 }= (1 +

*g*

*)*

_{t}*N*

*(4)*

_{t }3.3 Government

Government taxes labor income at rates *τ** _{w}*, and uses the revenues to finance an exogenously given amount of government consumption expenditures

*G*, which I assume is a fixed percentage of total output

*Y*

*. Both tax rates are determined endogenously*

_{t}in this model.

*G** _{t }*=

*τ*

*X*

_{w}

^{R}*i*=1

^{ }Y *i−*1

*j*=0

*ψ*_{j}* *

!

*w*_{i,t}_{+}_{i−}_{1}*N** _{i,t }*(5)

The government also runs a pay-as-you-go social security program that is financed entirely by a payroll tax *τ** _{ss}*.

X^{T}* *

*i*=*R*+1

^{ }Y *i−*1

*j*=0

!

*ψ*_{j}* *

*SS*_{t}*N** _{i,t }*=

*τ*

*X*

_{ss}

^{R}*i*=1

^{ }Y *i−*1

*j*=0

*ψ*_{j}* *

!

*w*_{i,t}_{+}_{i−}_{1}*N** _{i,t }*(6)

3.4 Households

In this model economy, the labor income of an individual is divided into two parts. Before retirement, the income of an individual at time *t *is given by the economy wide wage ratio *w** _{t}*. Once retired, he receives constant social security income

*SS*

_{t}*every year, which is equal to a fixed share of his wage ratio at the age of retirement,*

*µ · w*

_{R,t}_{+}

_{R−}_{1}. The after-tax income,

*e*

_{i,t}_{+}

_{i−}_{1}, of an individual is given by:

*e*_{i,t}_{+}_{i−}_{1 }=

^{} ^{}

(1 *− τ*_{w }*− τ** _{ss}*)

*w*

_{i,t}_{+}

_{i−}_{1 }

*i ≤ R SS*

_{t }*i > R*

(7)

where *i *= 1*, *2*, …*16, *R *= 8 is the official retirement age, *e*_{i,t}_{+}_{i−}_{1 }is the after-tax income of an individual of age *i *in time period *t *+ *i − *1, and *w*_{i,t}_{+}_{i−}_{1 }is the labor income of an individual of age *i *in time period *t *+ *i − *1.

The intertemporal budget constraint of an individual is given by: *k*_{i}_{+1}_{,t}_{+}* _{i }*+

*c*

_{i,t}_{+}

_{i−}_{1 }= (1 +

*r*

_{t}_{+}

_{i−}_{1})

*k*

_{i,t}_{+}

_{i−}_{1 }+

*e*

_{i,t}_{+}

_{i−}_{1 }(8)

where *k*_{i,t}_{+}_{i−}_{1 }is the individual wealth of an individual of age *i *in time period *t*+*i−*1, *c*_{i,t}_{+}_{i−}_{1 }is the consumption of an individual of age *i *in time period *t *+ *i − *1, and *r*_{t}_{i}_{−}_{1 }is the interest rate in time period *t *+ *i − *1.

9

Individuals do not care about their descendants or predecessors, which means they do not leave bequests to their children or provide financial support to their parents. Therefore, individuals face addition constraints given by:

*k*_{1}* _{,t }*=

*k*

_{T }_{+1}

*= 0 (9)*

_{,t }Individuals obtain utility from their own life-time consumption only and their utility function is characterized by constant relative risk aversion (CRRA). Therefore, the household’s maximization problem can be described as:

max

*{k*_{i,t}_{+}_{i−}_{1}*} *

X^{T}* i*=1

*β*^{i−}^{1}

^{ }Y *i−*1

*j*=0

*ψ*_{j}* *

^{!}*c*^{1}^{−γ}* *

*i,t*+*i−*1

1 *− γ *

subject to Equation 7, 8 and 9, where *β *is the subject discount factor, and *γ *is the degree of relative risk aversion.

3.5 Characteristics in equilibrium

In equilibrium, the households’ optimization problem is solved and all markets are cleared.

For labor market, assume one generation of individuals are endowed with one unit of labor, the labor supply in each period is given by:

*L*_{t }_{=}X^{R}* i*=1

^{ }Y *i−*1

*j*=0

*ψ*_{j}* *

!

*N*_{i,t−i}_{+1 }(10)

which means that the aggregate labor supply is equivalent to the sum of the working-age population.

For capital market, the aggregate capital in the economy is simply the sum of the individuals’ wealth:

*K*_{t }_{=}X^{T}* i*=1

^{ }Y *i−*1

*j*=0

*ψ*_{j}* *

!

*k*_{i,t}_{+}_{i−}_{1}*N** _{i,t }*(11)

For the goods market, the aggregate demand should be equal to the supply: *Y** _{t }*=

*C*

*+*

_{t }*I*

*+*

_{t }*G*

*(12)*

_{t }where the aggregate consumption *C*_{t }_{=}P^{T}_{i}_{=1 }^{ }Q_{i−}_{1}

_{j}_{=0 }*ψ*_{j}* *

gate investment *I** _{t }*=

*K*

_{t }*−*(1

*− δ*)

*K*

_{t−}_{1}.

10

*c*_{i,t}_{+}_{i−}_{1}*N** _{i,t}*, and the aggre

4 MATLAB Implementation

The procedures of solving for steady state basically follow hints on Problem Set 3. Step 1

Based on the individual’s optimization problem, I set up the Lagrangian equation to solve for the Euler Equation and derive the optimal path of individual wealth from it^{6}. The optimal path of individual wealth (Equation 16) is suitable for backward shooting algorithms.

Step 2

As is shown in Equation 17, the coefficients *α*^{1}*, α*^{2}*, α*^{3}*, α*^{4}in the optimal path need to be decided in addition to solving for the individual’s problem. The steady state interest rate and tax rates are determined endogenously in this model, and all of them depend on the aggregate capital in the equilibrium. Thus, I provide a guess of aggregate capital *K** _{t }*to obtain the coefficients and prepare for the backward shooting.

Step 3

The transversality condition in Equation 9 states that individuals leave no wealth upon their deaths (*k*_{T }_{+1}* _{,t }*= 0). By providing another guess of the second to last individual wealth

*k*

*, together with the wage ratio one period earlier, I get the value of third to last individual wealth*

_{T,t}*k*

_{T −}_{1}

*. Repeat this process until the first period of individual wealth*

_{,t}*k*

_{1}

*. If*

_{,t}*k*

_{1}

*is not approximately equal to zero*

_{,t }^{7}, I will update the initial guess of

*k*

*and repeat the above process so that eventually*

_{T,t }*k*

_{1}

*is close to zero*

_{,t }^{8}. These iterations can be achieved using

*while*and

*for*loops in MATLAB.

Step 4

The capital market clearing condition (Equation 11) requires that the sum of individual wealth should equal the aggregate capital in the economy. If the

^{6}The detailed derivation is shown in Appendix A.

^{7}Throughout this model, I choose 1 *× *10^{−}^{4 }as the tolerance level.

^{8}The update algorithm follows the hint on Problem Set 3.

*k*^{i}^{+1}

* _{T ,t }*=

*k*

^{i}

_{T ,t }*−*0

*.*01

*· k*

*1*

^{i}*,t*

where *i *is the number of iterations. If the initial guess of *k*^{i}* _{T ,t }*results in negative capital

*k*

^{i}_{1}

*in period 1, this rule will automatically choose*

_{,t }*k*

^{i}^{+1}

_{T ,t }*> k*^{i}* _{T ,t }*on the

*i*+ 1 iteration. Similarly, if

*k*

^{i}_{1}

*is*

_{,t }positive, this rule will give a smaller initial guess of *k*^{i}^{+1}

* _{T ,t }*.

11

the difference between the guess of *K** _{t}*in Step 2 and the sum of individual wealth obtained in Step 3 is greater than the tolerance level, the guess of aggregate capital

*K*

*is updated and Step 2 and 3 is repeated. The iteration will not stop until these two values converge. I use the fsolve*

_{t}*function in this step to solve for the equilibrium aggregate capital.*

12

5 Calibration

I calibrate the model so that the initial steady state matches the survival probability and old-age dependency ratio ^{9}in China in 2010s and the final steady state in the end of this century (around Year 2100). Most of the parameters are calibrated _{according to }^{˙Imrohoroglu} and Zhao (2018) work. Table 1 summarizes the main calibration used in the model.

Parameter Description Value

*β *Subject discount factor 1.01

*γ *Relative risk aversion 3

*δ *Capital depreciation rate 0.1

*θ *Capital output share 0.3

*µ *Benchmark pension dependency ratio 0.15

*T *Time horizon 16

*R *Benchmark retirement period 8

*A *TFP Factor 1

*q *Government expenditure to total output 0.16

*g*_{1 }Benchmark population growth rate at Stage 1 0.4%

*g*_{2 }Benchmark population growth rate at Stage 2 -0.5%

Table 1: Calibration

5.1 Demographics

As is mentioned in Section 3, the model period is 5 years and there are *T *= 16 periods in this economy. Individuals enter the economy at the age of 20 (*t *= 1) and retire at the age of 60 (*t *= *R*) in benchmark scenarios.

China is expected to experience dramatic demographic changes over the next several decades. In this thesis, I choose to model these changes in two stages, which corresponds to DESA (2019) projection. The first stage ranges from Year 2020 to 2030, when China’s population reaches its summit. The second stage is from Year 2031 to 2100, when the population declines gradually.

During 2010-2020, the population growth rate of China remains 0.5% on av ^{9}Hereafter the old-age dependency ratio is defined as the retired population (*t *= *R *+ 1*, R *+ 2*, …, T*) over the working population (*t *= 1*, *2*, …, R*).

13

erage^{10}. Therefore, I choose Year 2020 as the initial steady state and the population growth rate during this period is calibrated to 0.4%, which reflects the mild growth momentum of China’s population. The population growth rate in the second stage is calibrated to -0.5%, which follows DESA (2019) projection and the final steady state is set in Year 2100.

Table 2 summarizes the mortality risk at five-year intervals, which is used to calibrate the demographic changes in this model^{11}.

_{Age}Survival Probability

Initial Steady State Final Steady State

20-24 0.9972 0.9993

25-29 0.9964 0.9991

30-34 0.9955 0.9988

35-39 0.9939 0.9983

40-44 0.9913 0.9974

45-49 0.9863 0.9958

50-54 0.9772 0.9931

55-59 0.9593 0.9884

60-64 0.9261 0.9794

65-69 0.8691 0.9619

70-74 0.7817 0.9300

75-79 0.6753 0.8751

80-84 0.5462 0.7818

85-89 0.4052 0.6449

90-94 0.2866 0.4820

95-99 0.1944 0.2709

100+ 0 0

Table 2: Survival Probability

The benchmark population projection in this model is shown in Figure 2. Both the total population and the share of the working population (Age 20-60) shrink over time.

^{10 Data} is taken from the Seventh National Population Census results released on May 11, 2021.

See: http://www.stats.gov.cn/english/PressRelease/202105/t20210510 1817185.html ^{11}Data is computed based on the Life Table published by DESA (2019). See: https://population

.un.org/wpp/Download/Standard/CSV/

14

Figure 2: Population projection during Year 2020-2100

5.2 Preference and technology

The value of relative risk aversion *γ *is assumed to be 3, which is commonly used in literature. The subject discount factor *β *for individuals is set to 1.01, which I _{follow }^{˙Imrohoroglu} and Zhao (2018) finding to account for the mortality risk faced by individuals.

The capital depreciation ratio *δ *is set to 0.1 according to Bai et al. (2006), and the capital output share is set to 0.3. For simplicity, I assume there is no technological progress and assign the TFP factor *A *to 1, which is constant across the whole time horizon.

5.3 Government

The Chinese government expenditure is, on average, 16% of the GDP from 2000 to 2019^{12}. Based on this information, I calibrate the government expenditure to total output (*q*) to 0.16.

I assume there is no government deficit in the model and the labor income tax ^{12 Data} is taken from the World Development Indicators (WDI) database published by the World Bank. See: https://databank.worldbank.org/source/world-development-indicators#.

15

revenues exactly balance the government expenditure. The labor income tax rate in steady state is determined endogenously in the model.

The pension replacement rate^{13 }is calibrated to be 0.15, which follows the same _{value in }^{˙Imrohoroglu} and Zhao (2018). The pay-as-you-go social welfare program is also assumed to be self-financing. That is to say, the payroll tax revenue levied on working population balances the pensions received by the retirees in the economy. Hence, the payroll tax rate in steady state is also determined endogenously.

5.4 Policy reforms

As is discussed in Section 1, there are two potential policies the Chinese government could consider to reform: postpone retirement age and relax birth control policy.

Figure 3: Population projection with Three-child policy during Year 2020-2100

In this thesis, I will present results for three reforms which aims to deal with the aging population: (1) an increase of retirement age from 60 to 65; (2) a relaxation of birth control policy from ”two-child” to ”three-child”, which I assume would improve the population growth rate in both stages; and (3) a combination of reforms (1) and (2).

^{13}The pension replacement rate is defined as the pension income over the average wage ratio. 16

The calibration of retirement age reform is plain, I just changed the retirement period *R *from 8 to 9. To account for the impact of birth control policy reform, there are two adjustments in the benchmark demographics prediction. Firstly, I increase the horizon of the first stage from Year 2020 to 2040, and decrease the length of the second period from Year 2041 to 2100. Then I change the benchmark population growth rate in both stages to reflect the impact of the three-child policy. The population growth rate is set to 0.6% at stage 1 and -0.2% at stage 2. The population projection with this reform is shown in Figure 3.

17

6 Results

6.1 Time paths of individual wealth and consumption

In this subsection, I present the time paths of individual wealth and consumption in two steady states under different policy reforms. The results from these reforms are shown in Figure 4.

Panel A: Benchmark without policy reforms Panel B: Retirement age reform Panel C: Birth control policy reform Panel D: Both policy reforms

Figure 4: Time paths of individual capital and consumption

As is shown in Figure 4, individuals accumulate wealth during their working periods and consume it until death in steady states. To maximize their utilities, they choose to smooth consumption over their life. Their consumption is hump shaped with the maximum consumption appearing at their retirement period.

In the benchmark scenario without policy reforms (Panel A), Compared with initial steady states in 2020, the individual wealth level in 2100 is higher in any

18

period of life, and the consumption level in 2100 is lower in the early stage but exceeds the 2020 level in later periods of life. With an older population, individuals in 2100 choose a lower consumption level and accumulate more wealth in their early periods of life to support their consumption level in retirement. This is in line with the view that an aging population would promote more savings and accelerate capital accumulation.

Panel B demonstrates the effects of retirement age reform on steady state in 2100. The postponement of the retirement age from 60 to 65 has allowed individuals to have longer working periods and shorter retirement horizons during their life. Since individuals can earn a higher labor income during working periods, this reform enables them to accumulate wealth at a slower pace. Besides, the longevity risk faced by individuals decreases due to the shorter retirement horizon. These effects may explain the fact that the consumption level after reform dominates benchmark scenarios throughout the entire horizon.

The time paths of individual wealth and consumption with birth control policy reform do not deviate a lot from benchmark scenarios (Panel C). Although the total population after reform stays at a higher level under three-child policy, the distribution of population across various age groups remains relatively stable. Individuals choose to follow the same pattern of saving and consumption once the effect of reform fades away in the long run.

The impact of both reforms on individual wealth and consumption in steady states is simply the combination of each reform (Panel D). As discussed above, the dominant impact is from the retirement age reform.

6.2 Properties of steady states

The results for the potential policy reforms are shown in Table 3. Column *Bench mark *presents the properties of initial steady state in 2020, and Column *No Reform *represents the final steady state without policy reforms but in 2100 demographics. The results for retirement age reform, birth control policy reform and a combination of these reforms are presented in Columns *Reform 1*, *Reform 2*, and *Reform 3*.

19

Benchmark No Reform Reform 1 Reform 2 Reform 3

*2020 Demographics two-child policy three-child policy *

Total population (*N*) 1.0000 0.7238 0.7238 0.9578 0.9578 Old-age dependency ratio 29.30% 60.82% 43.39% 59.58% 42.49% Payroll tax rate (*τ** _{SS}*) 4.39% 9.12% 6.51% 8.94% 6.37% Aggregate output (

*Y*) 0.8046 0.5284 0.5676 0.7024 0.7535 Aggregate capital (

*K*) 0.8825 0.7686 0.7465 1.0140 0.9834 Aggregate labor (

*L*) 0.7734 0.4500 0.5048 0.6002 0.6722

*K/L*ratio 1.1411 1.7078 1.4789 1.6894 1.4629 Return to capital (

*r*) 17.35% 10.63% 12.81% 10.78% 12.99% Wage (

*w*) 0.7283 0.8219 0.7872 0.8193 0.7846 Pension income (

*SS*) 0.1092 0.1233 0.1181 0.1229 0.1177

*Note: The total population in 2020 is calibrated to 1.*

Table 3: Reform evaluation

In the model, the Chinese demographics has undergone dramatic changes from 2020 to 2100. The total population shrinks by about 30% (from 1 to 0.7238) under current two-child policy, and the old-age dependency ratio more than doubles in the No Reform* *scenario. The increase in retirement age of *Reform 1 *partially alleviates the rising old-age dependency ratio (from 60.82% to 43.39%). Compared with retirement age reform, the three-child policy succeeds in preventing the decline of the total population in the long run, but it fails to deal with the aging population. The old-age dependency ratio in *Reform 2 *is as high as the *No Reform *scenario. *Reform 3 *presents the best outcome in both maintaining population scale and deferring the rapid growth of old-age dependency ratio.

The aging population has imposed a great challenge to the pension system. Under *No Reform *scenario, the payroll tax rate *τ** _{SS }*increases from 4.39% to 9.12% to balance the pension expenditures. The rise of payroll tax rate will impact individuals’ disposable income negatively. As is shown in Equation 6, holding other parameters fixed, the payroll tax rate is mainly influenced by old-age dependency ratio. Therefore, postponing retirement age will be more effective in curbing the rising payroll tax rate than birth control policy reform (6.51% and 6.37% under different birth control policies), as it results in a lower old-age dependency ratio.

Due to the negative demographic trend discussed above, the aggregate labor in 20

*No Reform *scenario witnesses dramatic decline in 2100. Without reforms on birth control policy, it is shown in the table that retirement age postponement alone has limited effect in offsetting the negative impact of the aging population on the labor force, as the aggregate labor in *Reform 1 *(0.5048) is slightly higher than the No Reform* *case (0.4500). In comparison, the introduction of a three-child policy is relatively effective to make up the labor force gap between two steady states. As Council (2013) discusses, fertility rate improvement is the long-term solution to the aging problem. This argument is supported in this model.

Compared with aggregate labor, the impact of the aging problem on capital stock is milder in 2100. As is shown in Figure 4 Panel A, the individual wealth level in 2100 dominates that in 2020 under *No Reform *scenario. People tend to accumulate more wealth to maintain a higher consumption level after retirement. From the macroeconomic perspective, this increase in individual capital stock offsets part of the negative impact of changing demographics, resulting in a mild decrease in aggregate capital under the No Reform* *scenario. However, with the implementation of the three-child policy, this positive feedback from individuals outweighs the aging population, and the aggregate capital in the economy (1.0140 and 0.9834) exceeds the benchmark scenario (0.8825).

The impact on aggregate output can be obtained by Cobb-Douglas production equation once the aggregate capital and labor supply are fixed. Since the capital output share is 0.3 in this model, labor supply carries more weight in determining the output. This is reflected in the result, as the aggregate output under two-child policy witnesses a large drop relative to the benchmark scenario, and the relaxation of birth control policy keeps the output from declining further. It should be noted that the TFP factor is assumed to be constant over time in the model. When considering technological progress, the aggregate output should be larger compared with current results.

According to the results, the implications of an aging population in return to capital and wage rate are dramatically different: the return to capital in 2100 experiences considerable drop under all scenarios, whereas the wage rate witnesses an increase regardless of reforms. As is shown in Equation 1 and 2, the value of return to capital and wage rate is mainly governed by *K/L *ratio. The relative changes in aggregate

21

capital and labor supply in different scenarios result in the variations in return to capital and wage rate. The result has demonstrated that the capital stock is more abundant relative to labor supply in 2100. Therefore, the return to capital decreases due to oversupply and wage rate increases thanks to the aging population.

22

7 Conclusion

It is widely anticipated that China will experience dramatic demographic changes in the coming decades of this century. However, the rapid aging process far exceeds people’s anticipation^{14}. At the end of this thesis-writing period, the Chinese government unexpectedly relaxed its birth control policy and allows couples to have up to three children^{15}, which underlines the gravity of the aging problem. It becomes even more important to understand the consequences of these changes.

In this thesis, I examine the implications of the potential reforms in retirement age and birth control policy in China using the OLG model. It is shown that without reforms, the capital stock, labor supply and aggregate output will experience dramatic declines in the coming years of this century. The impact on return to capital and wage rate differs due to the changes in *K/L *ratio. The postponement of retirement age proves effective in delaying the rise of old-age dependency ratio, while birth control policy reform produces better results in coping with the aging population in the long run.

^{14}The latest Population Census shows that there are only 12 million babies born in 2020, continuing the descent to a near six-decade low. See: http://www.stats.gov.cn/english/

Press Release/202105/t20210510 1817185.html

^{15}On 31 May, 2021, the Chinese government lifted the cap on births and married Chinese couples are allowed to have up to three children. See: http://english.www.gov.cn/statecouncil/ ministries/202106/01/content WS60b61ab7c6d0df57f98da86e.html

Why China is the FinTech capital of the world?

8 References

Auerbach, A. J. and Kotlikoff, L. J. (1987). *Dynamic fiscal policy*. Cambridge Univ. Press, Cambridge.

Bai, C.-E., Hsieh, C.-T., and Qian, Y. (2006). The return to capital in china. Technical report, National Bureau of Economic Research.

Bell, D. N. and Rutherford, A. C. (2013). Older workers and working time. *The journal of the economics of aging*, 1:28–34.

Bernheim, B. D., Skinner, J., and Weinberg, S. (2001). What accounts for the variation in retirement wealth among us households? *American Economic Review*, 91(4):832–857.

Conesa, J. C. and Kehoe, T. J. (2018). An introduction to the macroeconomics of aging. *The Journal of the Economics of Aging*, 11:1–5.

Why China is the FinTech capital of the world?

Council, N. R. (2013). *Aging and the Macroeconomy : Long-Term Implications of an Older Population. *National Academies Press, Washington, D.C.

Davies, R. B. and Reed, R. R. (2006). Population aging, foreign direct investment, and tax competition. *University of Oregon Economics Department working papers*.

DESA, U. (2019). World population prospects 2019. United nations. department of economic and social affairs. *World Population Prospects 2019*.

Diamond, P. A. (1965). National debt in a neoclassical growth model. *The American Economic Review*, 55(5):1126–1150.

Heijdra, B. J. and Romp, W. E. (2009). Retirement, pensions, and aging. *Journal of Public Economics*, 93(3-4):586–604.

Hsu, M., Liao, P.-J., and Zhao, M. (2018). Demographic change and long-term growth in china: Past developments and the future challenge of aging. *Review of Development Economics*, 22(3):928–952.

^{˙Imrohoroglu}, A. and Zhao, K. (2018). Intergenerational transfers and china’s social security reform. *The Journal of the Economics of Ageing*, 11:62–70.

24

Lee, S.-H. and Mason, A. (2007). Who gains from the demographic dividend? forecasting income by age. *International journal of forecasting*, 23(4):603–619.

Lee, S.-H., Mason, A. W., and Park, D. (2011). Why does population aging matter so much for asia? population aging, economic growth, and economic security in asia. *Population Aging, Economic Growth, and Economic Security in Asia (October 1, 2011). Asian Development Bank Economics Working Paper Series*, (284).

Modigliani, F. and Cao, S. L. (2004). The Chinese saving puzzle and the life-cycle hypothesis. *Journal of economic literature*, 42(1):145–170.

OECD (2019). Aging and employment policies—statistics on average effective age of retirement.

Samuelson, P. A. (1958). An exact consumption-loan model of interest with or without the social contrivance of money. *Journal of political economy*, 66(6):467– 482.

Teixeira, A. A., Renuga Nagarajan, N., and Silva, S. T. (2017). The impact of aging and the speed of aging on the economic growth of least developed, emerging and developed countries, 1990–2013. *Review of Development Economics*, 21(3):909– 934.

25

A Appendix

To solve the household optimization problem, Lagrangian method can be used:

_{L }_{=}X^{T}* i*=1

*β*^{i−}^{1}

^{ }Y *i−*1

*j*=0

^{!}*c*^{1}^{−γ}* *

*i,t*+*i−*1

_{1 }* _{− γ}*+

*λ*

_{t}_{+}

_{i−}_{1}[

*k*

_{i}_{+1}

_{,t}_{+}

*+*

_{i}*c*

_{i,t}_{+}

_{i−}_{1}

*−*(1+

*r*

_{t}_{+}

_{i−}_{1})

*k*

_{i,t}_{+}

_{i−}_{1}

*−e*

_{i,t}_{+}

_{i−}_{1}]

*ψ*

_{j}

where *λ *is the Lagrange multiplier. The first-order conditions can be obtained by:

*∂L *

*∂c*_{i,t}_{+}_{i−}_{1}= *β*^{i−}^{1} *∂L *

^{ }Y *i−*1

*j*=0

!

*c*^{−γ}* *

*ψ*_{j}* *

_{i,t}_{+}_{i−}_{1 }+ *λ*_{t}_{+}_{i−}_{1 }= 0 (13)

*∂k*_{i}_{+1}_{,t}_{+}* _{i}*=

*λ*

_{t}_{+}

_{i−}_{1 }

*− λ*

_{t}_{+}

*(1 +*

_{i}*r*

_{t}_{+}

*) = 0 (14) From Equations 13 and 14, Euler Equation can be derived as:*

_{i}*c*_{i}_{+1}_{,t}_{+}* _{i }*= [

*βψ*

*(1 +*

_{i}*r*

_{t}_{+}

*)]*

_{i}^{1}

^{γ }*c*

_{i,t}_{+}

_{i−}_{1 }(15)

Using intertemporal budget constraint 8 to solve for the consumption and substitute into the Euler Equation 15. This will give the law of motion for individual’s wealth:

*k*_{i,t}_{+}_{i−}_{1 }= *α*^{1}_{t}_{+}_{i−}_{1}*k*_{i}_{+1}_{,t}_{+}* _{i }*+

*α*

^{2}

_{t}_{+}

_{i−}_{1}

*k*

_{i}_{+2}

_{,t}_{+}

_{i}_{+1 }+

*α*

^{3}

_{t}_{+}

_{i−}_{1}

*e*

_{i,t}_{+}

_{i−}_{1 }+

*α*

^{4}

_{t}_{+}

_{i−}_{1}

*e*

_{i}_{+1}

_{,t}_{+}

*(16)*

_{i }where

*α*^{1}_{t}_{+}_{i−}_{1 }_{=}1

1 + *r*_{t}_{+}_{i−}_{1}_{+}1

* _{i}*(1 +

*r*

_{t}_{+}

*)*

_{i}^{(1}

^{−γ}^{)}

*(1 +*

^{/γ}*r*

_{t}_{+}

_{i−}_{1}

_{)}(17)

*β*^{1}^{/γ}*ψ*^{1}^{/γ}* *

*α*^{2}_{t}_{+}_{i−}_{1 }_{= }* _{−}*1

* _{i}*(1 +

*r*

_{t}_{+}

*)*

_{i}^{1}

*(1 +*

^{/γ}*r*

_{t}_{+}

_{i−}_{1}

_{)}(18)

*β*^{1}^{/γ}*ψ*^{1}^{/γ}* *

*α*^{3}_{t}_{+}_{i−}_{1 }_{= }* _{−}*1

*− τ*

_{w }*− τ*

_{ss}

1 + *r*_{t}_{+}_{i−}_{1}(19)

*α*^{4}_{t}_{+}_{i−}_{1 }_{=}1 *− τ*_{w }*− τ*_{ss}* *