Location:Home > Achievements > Research Reports > Details

Ni Pengfei, Li Qingbin, and Li Chao:Spatial Heterogeneity of Urban Happiness and its Influencing Factors in China

发表于 lichao
Abstract: Based on the data of prefecture-level and above cities, this paper examines the spatial distribution of well-being by applying the spatial econometrics. The results of the study suggest thatChina’s urban well-being has a typical “club” characteristic and noticeable regional difference. As a relatively long-term and stable indicator, well-being is affected by variables of early conditions, urban per capita savings and public service, whereas inflexible indicators like urban per capita GDP, infrastructure and urban characteristic has little influence on the spatial heterogeneity of well-being. Attention should also be paid to the regional effect of the above affecting factors. In accelerating the construction of “happy cities”, it is particularly important to establish a “tilted and flat” urban system.

Key Words: Urban well-being, spatial heterogeneity, affecting factors, skew and flat

JEL Classifications:R13

 

1. Introduction

Nowadays,Chinahas gradually advanced from a low- and middle-income country to a moderate-income one whereas the transformation of socioeconomic development has entered a stage of overcoming tough problems. Against the background of various social problems, the whole society has been keeping questioning something like “Are we happy”. There are some relevant studies, but few have provided clear answers to the questions. Indeed, the issue requires comprehensive study from various aspects, of which “city” is a key analysis object. The primary reason is that the number and size of Chinese cities have been growing due to rapid urbanization. According to the data from the sixth nationwide census starting at the standard time—12 am on November 1, 2010,China’s urbanization rate has reached 49.68 percent if calculated by population ratio. Besides, we can predict from the development mode and experience of developed countries thatChinawill further accelerate urbanization in the following twenty years, which straightforwardly explains the importance of “city” as a special economy (region). Second, in the theory of urban economics, city is not only a center of various activities, but also an economic pattern with certain amount of cohesion (economies of agglomeration, economies of scale). Undoubtedly, various urban facilities and activities will play a more important role in people’s happiness. However, in the literatures we have found so far on Chinese people’s happiness, there is no study that takes city as an analysis unit. So this paper aims to fill in the blank and add a crucial aspect to happiness-related questions.

The data in this paper is mainly cited from the survey of resident well-being in nationwide prefecture-level and above cities conducted in 2010 by City and Competitiveness Research Center of Chinese Academyof Social Sciences. In this study, we take each city as an analysis object, whose average score of well-being and happiness index are defined as key variables. Data of variables is cited from China City Statistical Yearbook. This paper focuses on three major questions: (1) the spatial distribution of urban well-being inChina; (2) factors that affect urban well-being and spatial disparities; (3) the corresponding development strategy for such spatial distribution and affecting factors. Answering these questions will help us to recognize the facts of urban well-being and corresponding development strategy, thus laying the foundation for achieving the development goal of “happy cities.”

The structure of this paper is as follows: Part two gives a brief review of the existing studies of well-being and relevant research achievements based on China’s case; Part three describes the data and methods used in this paper; the fourth part focuses on features of spatial distribution by describing regional division and spatial distribution map; Part four examines possible affecting factors by applying spatial econometrics while taking into account regional effect; the last part suggests strategies based on empirical studies.

2. Review

There are a wide range of research topics in the studies of Subjective Well-being (SWB), which can be seen from the structure of Veenhoven, R’s “world database of happiness.” Among these topics, two closely related to the study in this paper are the measurement of well-being and its affecting factors.

The most common way to measure well-being is measuring the degree of happiness reported by individuals in a questionnaire. Some opponents argue that people are likely to provide false information in their report and the comparison is invalid given that the standard of happiness varies among different individuals, whereas supporters argue that the subjective experience of happiness is fundamentally effective for such self-report way, despite some deviation. Studies suggest that such subjective data can reflect the degree of people’s satisfaction with life, which is the most effective way of gaining subjective well-being (Lou 2009).

There are various factors that affect happiness, of which the most basic one is income (the level of economic development) versus happiness. In economics, individual happiness is measured by a person’s “utility”. The bigger the utility is, the happier the person is. The determining factors are the variables in utility function, of which the wealth or income level is regarded as the most important one. The higher the income is, the bigger the utility is, and the happier the person is. Such trend is most obvious in classical studies of economics and has led many countries and regions to pursue GDP—the indicator of material wealth—as the goal of happiness. This, however, has resulted in the paradox of happiness (Easterlin 1974): the residents’ happiness level has not improved with economic growth. So people started to verify and explain the paradox from various angles. There are basically two ways. On one hand, scholars try to explain the income paradox from the physical nature of micro individual happiness and argue that (1) the initial well-being will gradually decline after one gets used to the wealth growth; (2) well-being is determined by individual personality and remains respectively at a stable value in a life cycle of different individuals (Lykken and Tellegen 1996). So the income growth plays a very little role in well-being (Bruni 2007) whereas the impact of various short-term conditions is greatly limited, and well-being will eventually return to the fixed value.[1] On the other hand, more affecting factors have been discovered, which, as scholars argue, can also greatly determine one’s happiness. Apart from the factor of absolute income, there are (1) relative income and income gap. The competition for wealth among people greatly affects happiness. So the income gap between a person and those around him or her plays as a key factor and some studies even argue that happiness has nothing to do with absolute income but only relative income matters (Luis and Becker 2007); (2) the previous history and expectation, also known as temporal comparison. The change of various indicators over the time will affect well-being (Chen 2008); (3) recessive factors such as health, affection between family members and interpersonal relationships. These variables related to daily life also play a big role; (4) other macro factors such as inflation, unemployment, government expenditure and environment (Lu, Wang 2010). In short, the introduction of diverse affecting factors helps people to further understand well-being.

In studies that take Chinaas the object, scholars also focus on the relations between income and well-being, and affecting factors of well-being. In theoretical studies, Tian and Yang (2006) build a standard model to explain the paradox of income and happiness by dividing the affecting factors into income, non-income factors and relative effect. In this paper, we build a standard economic model by adding “omitted variable” and “aspiration theory” to get a critical income level. In empirical studies, most of them focus on clarifying the reality of Chinese people’s well-being and affecting factors. Some typical studies include the following ones. Luo (2006, 2009) examine factors that affect Chinese residents’ well-being by applying the nationwide urban and rural survey data, with a focus on the analysis of urban and rural division, employment, and absolute versus relative incomes, whereas Guan (2010) questions the conclusion reached by Luo (2006, 2009) and bases the study on sample data of Beijing, Shanghai and Guangdong from the 2008 survey by CFPS (Chinese Family Panel Studies) of Peking University. The study finds that absolute income plays an inconspicuous role whereas relative income has a remarkable positive impact on well-being. Zhu and Yang (2009) examine Chinese people’s well-being during 1990 to 2006 by applying world value survey data and other survey data. They find that the income in 1995, 2001 and 2006 has a positive correlation with happiness whereas the income in 1990 has a negative correlation with happiness. Xing (2011) systematically examines factors that affect urban residents’ well-being by applying open data and urban survey data, with a focus on income. Other similar studies include the one by Qi and Zhou (2010) that bases on the data of CHNS (China Health Nutrition Survey) and the study by Wu and Chen (2010) that bases on the 2009 urban survey data of ZhejiangProvince. In general, domestic studies are much the same as each other and the author has not found studies that take city as a key analysis object so far in the available literatures. So the basic task of this paper is to explore the spatial distribution of China’s well-being and affecting factors in the aspect of city.[2]

3. Description of Survey Data

The data cited in this paper comes from two sources. One is the survey of residents’ well-being in nationwide prefecture-level and above cities conducted in 2010 by City and Competitiveness Research Center of Chinese Academyof Social Sciences. The other source is the data of corresponding years in China City Statistical Yearbook, which is organized according to our research purpose (details omitted). Here is a brief explanation of the survey of well-being. It strictly follows standard procedures and is entrusted with a complete questionnaire to a professional survey company. The survey is conducted in the format of a large sample and a multi-dimensional questionnaire, which covers a total of 294 prefecture-level and above cities. It randomly picks up urban residents as the survey objects in the sample cities and asks them questions in the questionnaire through phone interviews, while the interviewer takes notes. The survey collects a number of 17,757 valid questionnaires after excluding the invalid ones. The questionnaire is made up by two parts of indicators, of which well-being is set as demonstrative indicator whereas various aspects of urban residents’ well-being such as income, living condition and health care are set as explanatory indicators. The question on well-being is: if the highest score of well-being is 100, how would you grade your state of well-being in 2010? It then proceeds to a retrospective survey of 2001, 2005 and 2009. The satisfaction of explanatory indicators is divided into five levels, namely, by selecting from 1 to 5, the satisfaction increases from low to high. Besides, the whole questionnaire is divided into a long version and a short version depending on the number of explanatory indicators; demonstrative indicator combines with a small number of explanatory indicators to form the short version, and combines with a large number of explanatory indicators to form the long version. The short version is applied to 259 ordinary cities and the long version to 35 key cities. There are also differences between the two versions in sample size, namely, the number of individuals who take the survey. By calculating the average value of well-being that takes city as a unit based on individual samples, the data of well-being in 294 cities can be described as follows:

Table 1: Descriptive Survey Statistics of China’s Urban Well-being















































Variable Sample Size Average Value Standard Deviation Minimum value Maximum value
The score of well-being in 2001 294 69.40 5.00 45.50 83.66
The score of well-being in 2005 294 70.52 4.19 52.90 81.70
The score of well-being in 2009 294 73.55 3.52 58.40 83.58
The score of well-being in 2010 294 76.06 3.49 66.57 89.36


Data resource: Survey data from China City and Competitiveness Research Center.

The average score of well-being in 294 cities in 2010 is 76.06. The highest score (89.36) is 1.34 times the lowest score (66.57) and the overall standard deviation is 3.49, indicating a small relative gap. The scores suggest that the urban residents inChinaas a whole are in a state of happiness and enjoy a high level of well-being, which, however, has not reached an “excellent and good” level but is at an average and above level. It is rare to reach such a level in a period of rapid economic development and sharp social transformation that sees various outstanding social problems. There has been a climbing trend over the four years: the average value of well-being has increased by seven points and people’s identification with happiness keeps growing. A look at the standard deviation indicating the dispersion among cities reveals that the extent of dispersion has been narrowing over the four years and the standard deviation that scores well-being has decreased from 5.00 in 2001 to 3.49 in 2010, indicating an overall trend of convergence in the well-being of urban residents. Such trend is indeed one of the features of spatial distribution. In the following part, we will study the issue in question based on the above data.

4. Feature of Spatial Distribution

4.1   Well-being Divided by Regions

We divide the cities in Chinese mainland into six regions—seven regions in total plus Taiwan. See Table 2 for the statistical description of the scores of well-being in each region. It suggests that the average score of the six cities in Taiwanexceeds 80, with relatively equal scores and a small dispersion,[3] which is obviously the best case among all the regions. Bohai coastal region also enjoys a high level of happiness and the average score of well-being is 79.48, which has a noticeable advantage over other mainland regions. But the dispersion among 30 sample cities is broader compared with the case of equalization inTaiwan. Apart from that, there are small disparities among the average value in other regions. The change over the four years suggests that a city’s location has a dynamic stable correlation with the distribution of well-being. Of the seven regions,Taiwan always tops the list whereas the northwestern region always scores lowest except in 2005. The basic characteristics of other regions also remain stable. Meanwhile, one of the changes is that the northeastern region replaces northwestern region in 2005 and scores lowest in well-being. Besides, a look at the change of standard deviation in each region reveals that most regions see a trend of convergence year by year (the standard deviation of some regions fluctuates in the middle years), except for the southeast region which sees an expanding dispersion. It suggests again thatChina’s urban well-being features a trend of spatial convergence.

Table 2: Different Regions and Urban Well-being












































































































Regions Number of Cities 2001 2005 2009 2010
Average value Standard deviation Average value Standard deviation Average value Standard deviation Average value Standard deviation
Northeast 34 65.51 6.45 68.05 5.10 72.34 4.42 75.61 3.75
Southeast 57 72.73 2.97 73.20 2.87 74.60 2.90 76.77 3.14
Bohai costal region 30 70.69 5.37 73.19 3.92 76.48 2.99 79.48 3.37
Taiwan 6 78.17 1.36 78.55 1.65 79.48 1.74 80.42 1.84
Northwest 39 66.52 4.67 68.21 3.72 71.83 3.23 74.23 3.42
Southwest 47 68.50 3.91 68.62 3.35 72.10 2.88 74.64 3.00
Central region 81 69.48 3.66 70.31 3.00 73.46 2.87 75.87 2.77

Note: Refer to p50 in The 2011 Report on China’s Urban Competitiveness for details of the cities in each region

4.2  Analysis of Spatial Correlation of Well-being

The spatial heterogeneity of well-being is largely determined by the spatial correlation of urban well-being. In Figure 1, it publishes the results of overall spatial autocorrelation test (Moran’s I) of the scores of well-being in sample cities in 2001 and 2010. According to Moran’s scatter plot, the regions under analysis can be divided into four parts, namely, H-H, L-H, L-L, H-L, which falls into Quadrant I, II, III and IV of the scatter plot respectively and indicates (1) a high value of itself and a high value of its neighboring quadrant, (2) a low value of itself and a high value of its neighboring quadrant, (3) a low value of itself and a low value of its neighboring quadrant, (4) a high value of itself and a low value of its neighboring quadrant. Comparing two endpoint years, we find that the value of Moran’s I—index ofChina’s urban well-being in 2001 is 0.1819 whereas that in 2010 is 0.2215, which indicates a remarkable increase of spatial autocorrelation. The scattered distribution of sample cities in these two figures more vividly illustrates such change. Compared with the left part of Figure 1, the right part of the figure has more scattered dots in Quadrant I and III, namely, the areas of H-H and L-L. Besides, the scattered dots symmetrically distribute along the two sides of fitted curve and the aggregation has a typical “club” feature. The growing spatial correlation requires us to take its impact on the result of regression seriously in our following analysis of affecting factors.

5. Affecting Factors

Let’s turn our attention to the factors that affect the scores of urban well-being. In order to determine the optimum strategy for urban development, we should further clarify the affecting factors of urban well-being. Some potential problems are the multiple collinear feature among variables and the neglect of spatial correlation in traditional regression. In this paper, we carefully solved these potential errors in our quantitative analysis.

5.1   Quantitative Model and Variable Selection

By evaluating the affecting factors in existing literatures and taking into consideration the particularity of cities, we draw from the practice of Zheng Siqi and etc. (2011) and propose the following quantitative model:

Explained variable is the average score of well-being in each city in 2010. As for explanatory variables, we assume —the average score of well-being in 2001—to be initial historical condition, and , and to be variables of income, urban characteristic, and infrastructure and public service. As the income variable,  mainly examines two indicators: per capita gross domestic product (PER_GDP) and per capita saving deposit of urban and rural residents (PER_SAVE) by the end of the year, which respectively measures the economic development and savings in each city. As for the variable of urban characteristic , we borrow the indicator of abundance of natural resources “NR” from Xu Kangning and Wang Jian (2006) and calculate the proportion of extractive industry employees in total local employees; Environment variable “ENV” is measured by two indicators, namely, centralized sewerage disposal rate (ENV1) and per capita green area (ENV2); “OPEN” is the extent of opening up to the outside world, which in this paper, is measured by the proportion of total value of import and export in GDP of each city based on conventional practice. LEVEL is the dymmy variable of the levels of cities. We assign the value “1” to independently administered municipal cities, sub-provincial cities and provincial capitals and “0” to other prefecture-level cities to measure the disparities of urban well-being in cities of different administrative levels. As for , we adopt two indicators, namely, per capita road area (PER_ROAD) and the satisfaction of urban residents with local traffic condition (E_TRAFFIC). Of the above indicators, the indexes of happiness (,) and satisfaction of urban residents with local traffic condition “E_TRAFFIC” come from the survey data and other indicators are cited from China City Statistical Yearbook 2010.

5.2   Test of Spatial Model and Result of Regression

Given that traditional quantitative method fails to take into account the spatial correlation between different regions, it suffers an obvious limitation in describing spatial distribution and spillover effect. So this paper attempts to analyze spatial regression by adopting Spatial Lag Model (SLM) and Spatial Error Model (SEM) to overcome the shortcomings. Apart from the previously mentioned Moran’s I, the test of spatial correlation can also be conducted by applying two Lagrange Multipliers LMERR and LMLAG, and Robust R-LMERR and R-LMLAG. Anselin and Florax (1995) proposed the following criteria: if neither results of LMLAG and LMERR are significant, keep the original result of OLS; if the statistical result of LMLAG is more significant than that of LMERR in the test of spatial correlation, so with R-LMLAG and R-LMERR, it can be confirmed that the suitable model is Spatial Lag Model; on the contrary, if the statistical result of LMERR is more siginificant than that of LMLAG, so with R-LMERR and R-LMLAG, it can be confirmed that Spatial Error Model is the suitable model. Based on the above procedure, we apply LM to examine Formula 1. See Table 3 for the results.

According to the criteria proposed by Anselin and Florax (1995), the fitting result of Spatial Lag Model (SLM) is more satisfying. For the convenience of a direct comparison of several estimation methods, this paper also lists the estimated results of Spatial Error Model (SEM) and Ordinary Least Square (OLS), as suggested in Table 4.

Table 3: Test and Diagnosis of Spatial Correlation


















































Test



MI/DF



Standard Value



Value of P



Moran's I (error)



0.128921



3.5607566



0.0003699***



Lagrange Multiplier (lag)



1



11.3384605



0.0007592***



Robust LM (lag)



1



3.1955781



0.0738377*



Lagrange Multiplier (error)



1



8.2105420



0.0041648***



Robust LM (error)



1



0.0676596



0.7947753



Lagrange Multiplier (SARMA)



2



11.4061201



0.0033357***




Note: ***indicates statistical significance at the rate of 1%, ** indicates statistical significance at the rate of 5%, * indicates statistical significance at the rate of 10%.

Table 4: Analysis of Factors of Spatial Heterogeneity of China’s Subjective Urban Well-being












































































































Estimated methodVariable & Test

Spatial Lag Model


(SLM)



Spatial Error Model


(SEM)



Ordinary Least Squares


(OLS)



Constant



28.763(0.000)



43.834(0.000)



42.559(0.000)



H2001



0.296(0.000)



0.301(0.000)



0.317(0.000)



PER_GDP



6.490e-006(0.464)



6.508e-006(0.473)



7.136e-006(0.444)



PER_SAVE



1.036e-005(0.009)



1.099e-005(0.008)



1.083e-005(0.010)



NR



1.455(0.435)



1.094(0.572)



1.773(0.366)



ENV1



0.008(0.299)



0.008(0.314)



0.011(0.154)



ENV2



-0.002(0.509)



-0.003(0.483)



-0.002(0.432)



OPEN



-0.099(0.877)



-0.032(0.963)



-0.04225167(0.950)



LEVEL



0.438(0.497)



0.473(0.465)



0.233(0.731)



PER_ROAD



-0.072(0.027)



-0.079(0.017)



-0.066(0.057)



E_TRAFFIC



3.229(0.000)



3.330(0.000)



3.273(0.000)



R2



0.348



0.343



0.312



Log likelihood



-607.913



-608.897



-612.99



AIC



1239.79



1239.83



1247.98



SC



1278.40



1281.94



1286.58



Likelihood Ratio Test



10.156(0.001)



8.186941(0.004)


 

Note: The value in the parenthesis is p. LIK, AIC, SC are the indicators that examine the fitness of the model. The larger the value of LIK is, the better the fitness is, whereas the smaller the value of AIC and SC is, the better the fitness is. Given that there are many explanatory variables in this paper, we examine Tolerance and Variance Inflation Factor (VIF). The results suggest that the VIF of all explanatory variables is below 5 whereas the tolerance is above 0.2, so there is no need to worry about multiple collinear problems. The test result of BP is also satisfying.

From the comparison of the estimated results of three models, we can find that most of the estimated coefficients in SLM see a decline in absolute value as opposed to OLS. To some extent, these variables gain explanatory capability through the spatial lag variables in their neighboring area. For spatial lag residual error, the statistic of Moran’s I is -0.0079, a figure almost approximate to zero, which suggests that the spatial lag variables in SLM have excluded the possibility of spatial autocorrelation. Meanwhile, the comparison of fitness among models reveals that SLM has an advantage over the other two estimation methods. So we are likely to reach inaccurate and even wrong conclusions owing to inappropriate models if we neglect the spatial correlation and estimate arbitrarily in our analysis of the affecting factors of urban well-being.

From the explanatory variables of SLM and the test results in Table 4, we find that the coefficient of H2001 has a significant positive value at the rate of 1 percent, which indicates an outstanding feature of territorialism and continuity in the score of urban well-being. That the present score of well-being is influenced remarkably by the historical counterpart suggests that the score is a long-term and slow indicator with a profound historical impact. Among the other factors that affect the index of urban well-being, the income variable has always been a hot subject of academic studies. The results of this paper suggest that the per capita GDP “PER_GDP” that measures economic development fails in the test of statistical significance whereas the per capita saving deposit “PER_SAVE” of urban and rural residents by the end of the year has a highly significant value. This suggests that the indicator of per capita GDP is not applicable to the discussion of well-being whereas the savings closely related to personal income level is the key factor that affects well-being, which requires us to shift the focus from increasing GDP to securing resident income. The variable of urban characteristic include the abundance of natural resources (NR), pleasant environment (ENV), the extent of opening up (OPEN) and the level of cities (LEVEL), all of which fail in the test of statistical significance at the rate of 10 percent. Such result suggests that a city’s resources, environment, opening up, as well as its size and administrative level has little influence on urban well-being in the analysis framework of spatial lag model. We argue that it is largely because these variables usually serve as intermediate ones in SLM. Of the variables of urban infrastructure and public service, the two indicators “PER_ROAD” and “E-TRAFFIC”, namely, urban per capita road area and urban residents’ satisfaction with local traffic condition suggest a highly significant relativity, which features weak negative correlation and strong positive correlation respectively. This reflects a key feature in public facility and service in China’s current stage of development, that is, large-scale investment in urban infrastructure is accompanied by the absence of efficient planning and management, which results in deviation from the principle of making the most and reasonable use of infrastructure from the very start. In general, the development under this category is still characterized by an extensive mode which ends up in a situation of more roads but more traffic jams. The conclusion also has an obvious significance in policy making. In order to achieve the goal of improving urban well-being, local governments should endeavor to improve public service by taking into account their own reality to make the most and reasonable use of existing and newly constructed infrastructure, which proves to be more effective than large-scale investment in urban infrastructure construction.

5.3   Regional Effect of Affecting Factors

The existing studies show that there could be big differences in factors that determine and affect well-being in different regions. So in order to further clarify the different impact of spatial heterogeneity on the well-being in different regions, we adopt the advanced multivariate method known as ESDA to examine the conditional scatter pot of significant variables in Table 4. [4]

The results of data fitting suggest that (see Figure 2, 3, 4, 5) there are tiny regional disparities in the regression slope of well-being and four explanatory variables. The slopes of each region in Figure 2 and 4 are much the same as each other, which suggests that historical well-being and public service has a positive impact on the immediate well-being. The conclusion is basically the same as our judgment of the whole sample. In Figure 3 and 5, there are frequent cases of opposite slopes between regions. Take the impact of income in each region on the immediate well-being for example, the fitting slope of income increase in the western region has a positive correlation with the immediate well-being whereas that in the southeast coastal region sees a negative correlation. Similarly, in the western region with relatively backward infrastructure, increasing investment in its infrastructure will greatly improve the region’s immediate well-being, in contrast to the case of the northeastern and southeastern regions with relatively advanced infrastructure. Generally speaking, given the objective reality of spatial heterogeneity, there are some small discrepancies between the results of data fitting in each region and the judgment of the whole sample. It is noteworthy that there would be differences in the precision of estimated slope correlation coefficients due to the unbalanced number of observation sties in each region. Given the large sample, the outlier of some regions will not affect our earlier conclusion. However, the regional effect of affecting factors of well-being suggests that governments of various levels should come up with different measures that target at the development of “happy cities” in different regions by setting different goals.

6. Conclusion and Suggestion for Policy-making

The study in this paper shows that the spatial distribution ofChina’s urban well-being is characterized by noticeable imbalance. In our discussion of the affecting factors, we find that historical data has a profound impact on a long-term stable indicator like well-being, while income-related indicators play a bigger role in affecting well-being than per capita GDP. The variable of urban characteristic has no impact on well-being in the framework of spatial lag model. And it is not necessarily the case that the more infrastructure the better, for people’s well-being depends more on the satisfaction with the management and service that accompany the infrastructure. Besides, the affecting factors also feature regional effect, resulting in different factors of the well-being in different regions and their varying impacts on these regions. These factors and their regional effects can largely explain the spatial disparities ofChina’s urban well-being.

Based on these study results, we argue that the government should come up with a long-term strategy on a macro level. The goal of the macro strategy is to ensure the healthy development ofChina’s urbanization while promoting the comprehensive, harmonious and sustainable economic and social development, thus achieving the goal of building “happy cities”. In this paper, we propose an overall strategy of establishing a “tilted and flat” urban system, in which “tilted” is reflected in spatial planning and regional urban development whereas “flat” is reflected in basic system and public service. The “tilted and flat” urban strategy is reflected in:

6.1   Spatial Planning and Tilts Strategy of Regional Development

First, the government should map out a moderately tilted spatial structure. The spatial correlation of urban well-being and its resulting spatial aggregation is the starting point of tilts strategy. Based on the planning of main functional region, the government should make and implement systematic plans of spatial layout of nationwide urban and rural areas and regional development. By taking into account historical background and other natural endowments, it should establish a moderately tilted spatial structure of nationwide urban and rural economic and social activities and future development to direct the reasonable migration and balanced distribution of population and industrial space, which will eventually lead to a network of spatial pattern with harmonious development among big-medium-small cities, small towns and residential areas, led and connected by many urban groups. Second, the government should avoid making a single set of standards and measures of happiness for diverse cases. Well-being is indeed an issue that involves different stages. Moreover, given the imbalanced spatial distribution ofChina’s urban well-being, we should avoid replacing the development mode in other stages with the mode of developed cities and setting a unified standard in different regions, as well as preventing cities from competing for some indicators of happiness in conflict with their own reality. In the present stage, tilts strategy usually means setting different standards. Third, the government should pay attention to and prioritize peripheral regions for aid. In planning and building cities and urban groups, it should make positive and efficient fiscal, financial and industrial policies while focusing on aiding the rural areas, non urban groups and the development in peripheral regions of urban groups, as well as increasing investment in infrastructure and public service so as to avoid “Merton Effect” and reduce the degree of regional inclination.

6.2 Flat Strategy of Basic System and Public Service

The standard and planned tilts strategy should not affect the equalization of basic system and public service. First, the government should reform the system and policy featuring urban-rural and regional divisions and establish step by step the population management system featuring the unification of urban and rural areas and the free and orderly migration of population. It should reform the land system and establish the same-right, same-price system of land use and management featuring the unification of urban and rural areas, as well as removing the administrative barriers, urban-rural division and barriers between the regions to establish a nationwide unified market that plays a basic role in resource allocation, allowing factors of production and product service to flow fully and freely among nationwide urban and rural areas, and different regions. Second, the government should establish a perfect network of infrastructure characterized by overall planning and service rather than an extensive system through large-scale investment, including building and improving the network of fast, efficient, convenient and cheap infrastructure such as nationwide traffic, communication and information, while actively improving the management and service of infrastructure to make more reasonable and efficient use of it. The goal is to achieve the fast, convenient, cheap and free flow of resources, population and information between regions, cities and urban and rural areas, thus allowing as many regions as possible especially rural and peripheral towns to join the network of infrastructure and win an opportunity to participate in nationwide and global competition and cooperation. Third, the government should promote the nationwide unification and equalization of public service. It should promote the unification and equalization in a larger scale in public service such as employment, social security, education and health care between different regions and urban and rural areas, reduce the degree of regional inclination and increase the mobility of population and labor forces while preventing the over aggregation of population to reduce the cases of “urban social problems” resulting from overcrowding and insufficient infrastructure and public service.

Happiness is fundamentally a subjective concept, whereas the basic factors behind it are objective. There are various factors that contribute to happiness. While we can do nothing about historical impact, it is the public’s expectation for the government to provide a sound public service system functioning as the objective base for people to gain happiness.

 

References:


[USA] Luigino Bruni and etc., Economics and Happiness, trans. by Fu Hongchun and etc., Shanghai People’s Publishing House, 2007.


Chen Huixiong,“Cross-disciplinary Deduction of Happiness Theory: Research Summary”, Anthology on Finance and Economics, Issue 1, 2008.


Fang Ying, Ji Kan, Zhao Yang, “Is There a ‘Resource Curse’ in China”, World Economy, Issue 4, 2011


Guan Hao, “Study of Impact of Income on Well-being: Absolute Level and Relative Status”, Nankai Economic Studies, Issue 5, 2010.


Huang Liqing, Xing Zhanjun, “Overseas Studies of Affecting Factors of Subjective Well-being”, Overseas Social Sciences, Issue 3, 2005.


Li Chao, Qin Chenglin, “Factor Endowment, Resource and Environmental Constraints on Spatial Distribution of Chinese Modern Industry”, Nankai Economic Studies, Issue 4, 2011.


Lou Lingli, “Development of Economic Studies of Subjective Well-being”, Development of Economic Studies, Issue 2, 2009.


Lu Yuanping, Wang Tao, “Review of Studies of Affecting Factors of Subjective Well-being”, Development of Economic Studies, Issue 5, 2010.


Luo Chuliang, “Absolute Income, Relative Income and Subjective Well-being”, Studies of Finance and Economics, Issue 11, 2009.


Ni Pengfei, The 2011 Report on Competitiveness of Chinese Cities, Social Sciences Academic Press, 2011.


Qi Shouwei, Zhou Shaofu, “Impact of Income, Health and Health Care on Well-being of the Elderly”, Public Management Journal, Issue 1, 2010.


Tian Guoqiang, Yan Liyan, “Answer to the Happiness-Income Puzzle”, Economic Studies, Issue 11, 2006.


Wu Limin, Chen Huixiong, “Model Building of Structure Equation of Income and Happiness”, China’s Rural Economy, Issue 11, 2010.


Wu Yuming, “Aggregation and Heterogeneity in Regional Economic Growth: Empirical Studies of spatial econometrics”, World Economic Papers, Issue 2, 2007.


Xing Zhanjun, “Study of Income and Well-being of Chinese Residents”, Studies of Social Sciences, Issue 1, 2011.


Xu Kangning, Wang Jian, “Study of Abundance of Natural Resources and Economic Development Level”, Economic Studies, Issue 1, 2006.


Zheng Siqi, Fu Yuming, Ren Rongrong, “Urban residents' willingness-to-pay for quality of life: evidence from housing cost changes and convergence”, World Economic Papers, Issue 2, 2011.


Zhu Jianfang, Yang Xiaolan, “Empirical Study of Income and Happiness in China’s Transformation Period”, Statistical Studies, Issue 4, 2009.










[1] Others argue that one’s happiness will fail to return to the “fixed value” after experiencing a certain impact. Readers interested in the topic can refer to Lou (2009)’s summary and other relevant literatures.




[2] The affecting factors in the studies that take city as an analysis unit are different from those in the studies of individual well-being, as suggested in the following results.




[3] The results are partly affected by the relatively small sample size, but they still indicate a small degree of dispersion.




[4] First, we take the longitude and latitude of scattered sites of each city as conditional variables by placing them respectively on horizontal and vertical axes to indicate the west-east and north-south directions. Based on geographical locations, we divide the whole sample into nine regions. Moreover, we take the average score of 2010 well-being as explained variable and the significant variables in Table 4 including historical well-being, income, public service and infrastructure as explanatory variables.