Sage Journals: Discover world-class research

Abstract

Taxi fare is related to the daily life of residents. Reasonable taxi fare not only meets the travel demand of residents but also improves the income of drivers and promotes employment. The spatial variation of taxi fares exists between different regions and cities. Previous studies on taxi fares have been conducted mostly in individual cities, and there has been no study on the spatial differentiation pattern of taxi fares on a national scale. Taking 336 cities across China as the research object, a multiple linear regression model of taxi fares was established by demonstrating the spatial variation pattern of taxi fares, the global differentiation index, spatial autocorrelation analysis, and kernel density analysis, etc. The significance of the study is to explore the law of spatial differentiation of taxi fares in China and to provide a stenographic record of taxi fare adjustment. The results show that: (1) the spatial variation of taxi fares across the country is significant, with the starting taxi fare range being between RMB 4 and RMB 14. (2) The global differentiation index of taxi fares is large, with two low-low clusters and three high-high clusters appearing spatially, and the results of the kernel density analysis surface a dispersion distribution centered on provincial capitals. (3) The divergence pattern of taxis nationwide is influenced by several factors. A multiple linear regression model is selected to establish a multiple linear regression model of urban disposable income per capita, regional GDP, urbanization rate, and urban population density, which shows that urban disposable income per capita has the greatest influence on taxi fares. The model shows that urban disposable income per capita has the greatest influence on taxi fares, and the fare of a 5 km taxi ride is 6.07. Taxi fares have a clear pattern of spatial differentiation in China and are most affected by urban disposable income per capita. Through this study, we can gain a deeper understanding of the variation in taxi fares across the country and provide data and theoretical support for the rationality of taxi fare adjustments.

Keywords

taxi fares spatial differentiation prefecture-level municipal units China

Introduction

Taxi services make a significant contribution to mobility in medium and large cities in almost every country in the world (Gallo, 2018), and taxis have become an important mode of transport for people going to work, shopping, and other travel activities (C. Hu & Thill, 2019). Taxis have many beneficial and desirable features, such as personalized, door-to-door, and demand response services (Wong et al., 2020). As an indispensable mode of transport in large cities, taxis can fill the gap between other modes of travel in terms of space and time (Huang et al., 2021; M. Li et al., 2017). The number of taxis in China has grown steadily since the 21st century, with positive growth in all years except 2006 and 2018, when there was a small decrease, with an average annual growth rate of 3.26%. In particular, the number of taxis in China grew rapidly in 2021, with a growth rate of 24.99% (Figure 1). Reasonable taxi fares meet the commuting needs of residents while improving driver incomes and increasing employment (Phiboonbanakit & Horanont, 2016). Therefore, it is necessary to conduct a study on taxi fares. Are there differences in taxi fares in different cities, and what factors influence such differences?

Figure 1.

Total number of taxis in China from 2000 to 2021.

Existing research on taxi fares has mainly focused on fare forecasting or improvement in a particular city. Studies on taxi fares include how to formulate a reasonable pricing strategy for taxis (Gallo, 2018; Gan et al., 2013; B. Li et al., 2022; T. Liu et al., 2019; H. Yang et al., 2010; Siyal et al., 2021), how to balance the supply and demand of urban taxis through taxi fares (Kamga et al., 2015; J. Wang et al., 2013; H. Yang et al., 2000), policy and regulatory issues on taxi fares (Hoover, 2022; Schaller, 1998; Wyman, 2017), and the relationship between taxi fares and services (Xu et al., 1999; H. Yang et al., 2002; T. Yang et al., 2014). For instance, W. Zhang et al. (2018) built a model to analyze how taxi drivers find customers when they get off the train outside the city center and to determine which internal and external factors influence their preferences (W. Zhang et al., 2018). Guo et al. (2021) found time-dependent pricing could avoid cross-subsidization among travelers in different periods based on a microeconomic model for the design of a time-dependent transit pricing scheme considering elastic and spatiotemporally distributed demand (Guo et al., 2021). A study has applied a mathematical model to the pricing of taxis in the city of Barcelona, and the optimum fare for the taxi sector is 0.68 euros/km and the optimum number of taxis is 19 taxis/km² (Grau & Estrada, 2019). Silveira-Santos et al. (2023) combined supervised and unsupervised techniques to predict online taxi fares in Madrid in comparison with taxi fares (Silveira-Santos et al., 2023). Wong and Szeto (2022) proposed a surcharge on taxi customers taking taxis during peak hours and/or traveling to congested areas as a way to address the acute shortage of taxi supply during peak hours (Wong & Szeto, 2022). The study found that fare cost is the most important factor driving transport mode preference when there are multiple options available and taxi drivers should focus their business on meeting the needs of these customers (Ulahannan & Birrell, 2022).

Research on taxis and spatial patterns has focused on the micro level within cities. Relevant topics of the study include energy consumption and pollutant emissions from taxis (Luo et al., 2017), the impact of urban-rural spatial structure on mobility networks (Nie et al., 2023), spatial-temporal characteristics and patterns of urban travel (H. Wang et al., 2019), Structure of the travel trajectory network (Y. Yang et al., 2018), and so on. The methodology of the study includes exploratory spatial data analysis (Deng & Ji, 2011), a multiscale geographically weighted regression model (Lyu et al., 2023), a spatial network-based Markov decision process approach (Zhou et al., 2018), and so on. Related application directions include mining the spatio-temporal patterns of urban residents (Mao et al., 2016), modeling the spatial and temporal distribution of the population (Rahimi et al., 2021), projections of urban transportation emissions (J. Liu et al., 2019), and so on. Taking Wuhan as an example, Tang et al. (2017) calculated the performance of taxi drivers based on their GPS trajectories for a week, classified taxi drivers into top drivers and ordinary drivers, and analyzed the spatial and temporal distribution and operational patterns of the top drivers (Tang et al., 2017). Using New York as an example, W. Zhang et al. (2020) explored the impact of the presence of an app-based taxi service on total daily taxi and street taxi ridership, and the impact of dynamic pricing on app-based hourly taxis (W. Zhang et al., 2020). Using Seoul as an example, Choi et al. (2023) measured the taxi pooling effect and its spatial and temporal patterns and proposed a method for identifying high-demand taxi pick-up and drop-off areas that can support a limited shared system (Choi et al., 2023). Z.-T. Li et al. (2023) processed taxi trajectory data from Chengdu and New York City to empirically and systematically analyze these urban travel networks based on network hierarchies, based on complex network theory (Z.-T. Li et al., 2023). Using the density of the gross domestic product (GDP), density of population, rate of urbanization, and access to transportation as evaluation indicators, Jin et al. (2020) rebalanced spatial protection and development of Wuhan city (Jin et al., 2020). In general, most of the existing research on taxis has focused on a particular city and no studies have examined the spatial pattern of taxi fares at a macro level.

Research on the spatial differentiation of taxi fares and their drivers can be based on location theory (North, 1955), supply and demand theory (H. Yang et al., 2002), and pricing theory (X. Wang et al., 2016). Among them, location theory emphasizes that any geospatial phenomenon is related to the location in which it is situated (Chan, 2001). At the municipal scale, taxi fares are closely related to the economic development of the city in which they are located, and taxi fares are generally higher in cities with better economic development (Cetin & Deakin, 2019). Supply and demand theory focuses on the game between supply and demand (Thathsarani et al., 2023). Taking taxi prices as an example, the supply side is the taxi company and the demand side is the passengers, and taxis have the attributes of a bilateral market. Passengers and drivers have complementary demands, and an increase in the number of passengers or a decrease in the number of taxi drivers will make the supply smaller than the demand, thus affecting taxi fares, and vice versa (J. Wang et al., 2013; H. Yang et al., 2002). Pricing theory is to make decisions about the price of a product under certain objectives, and the commonly used pricing methods are cost-oriented pricing method, market-oriented pricing method, and competition-oriented pricing method. The cost-oriented pricing method prices taxis based on their costs (including vehicle depreciation, fuel costs, management fees, etc.) plus a certain amount of profit. Market-oriented pricing is a consumer demand-centered pricing method that prices taxis based on consumer demand, willingness to demand, and cost of living. Competition-oriented pricing method is a pricing method centered on the pricing of competitors, and the main competitors of taxis include internet taxis, buses, and subways (Gómez-Lobo et al., 2022; He et al., 2018). This paper takes the city as the basic research unit. It focuses on the law of divergence and influencing factors of taxi fares among different cities, so it is more suitable to use the location theory as the theoretical framework. Under this theory, the taxi fares in different cities are different, which mainly depends on the different locations in which the cities are located. The more specific manifestations of city location are total GDP, population size, city size, distance from the sea, capital city, provincial capitals, and so on.

The geographic environment is a whole, and the elements that make up the geographic environment influence and constrain each other, together shaping the overall characteristics of the geographic environment. Different geographic environments also have regional differences, and the law of such differences is the law of spatial differentiation of the geographic environment (Jiang, 2015). The study of the law of differentiation between physical and human geography is one of the important directions of geography, as the totality and variation of the geographic environment are its basic characteristics. Human geography studies on the law of differentiation mainly include: the law of spatial differentiation between urban and rural development (Chen et al., 2019; Feng et al., 2023; Hou et al., 2022; Y. Yang et al., 2021), the law of spatial differentiation between tourist attractions and leisure places (Kang et al., 2022; Liao et al., 2022; P. Wang et al., 2022; Xiao et al., 2023), the law of spatial differentiation between transportation development and accessibility (D. Wang et al., 2022; J. Wang et al., 2020; Zhao et al., 2018), and the law of spatial differentiation between economic development and price levels (Y. Liu & Fang, 2022; Wu et al., 2020; Xing & Ye, 2023; Yu et al., 2023; L. Zhang et al., 2022). The obvious pattern of divergence in price levels of different commodities in different parts of the country has become an important topic of research in human geography and regional economics (S. Wang et al., 2016). In general, existing studies on taxi fares focus on the micro level, and take a city as an example to construct a model to calculate a reasonable taxi fare. Fewer studies have analyzed the spatial variation of taxi fares on a large scale. Reasonable taxi fares play an important role in meeting residents’ travel needs, increasing drivers’ income, and promoting employment. Therefore, this study takes Chinese cities as an example and intends to address two main questions. First, what is the spatial pattern of taxi fares in China, which cities have higher and lower taxi fares, and have agglomerations formed? Second, what factors are associated with the formation of this spatial pattern, and can a model of taxi fares be derived from the influencing factors?

Materials and Methods

Study Area and Data Sources

The 333 prefecture-level cities (including prefecture-level cities, regions, autonomous prefectures, and leagues) and four municipalities directly under the Central Government, excluding Sansha City in Hainan Province, have a total of 336 basic research units, excluding Hong Kong, Macao, and Taiwan.

Data on taxi fares were mainly obtained from taxi fare adjustment scheme announcements published on the Internet by the development and reform commission, municipal government, and municipal transport commission of each prefecture-level city. The taxi fare policy varies from place to place and mainly consists of the starting price, base price per kilometer, emptying or idling fee, waiting or low-speed fee, nighttime fare increase, and special compensation fee. Waiting or low-speed charges are mainly caused by passengers requesting to wait, traffic jams, waiting for red lights, etc. They are generally free within 5 min, assuming ideal conditions where passengers do not request to wait there are no long traffic jams, and the total time spent waiting for red lights does not exceed 5 min, that is, there is no waiting charge. This study assumes that taxis all travel during the day, that is, there is no night-time fare increase. In some cities, a special compensation charge is added during the Chinese New Year period, usually at a rate of 5 to 10 RMB on top of the original fare, assuming it is not during the Chinese New Year period, that is, there is no special compensation charge. For the sake of consistency and ease of calculation, the taxi fares calculated in this article include only the starting price, base price per kilometer, and emptying charge.

Other data is obtained from provincial and municipal statistical yearbooks and bulletins.

Methods and Formulas

The research methodology is divided into three main parts, including spatial divergence phenomena, spatial divergence characteristics, and modeling. The first two of these characterize the spatial divergence patterns and characteristics of taxi fares in China and demonstrate and analyze the spatial divergence of prices, followed by the development of a taxi fare model to measure the factors influencing taxi fare divergence and the magnitude of its intensity.

The Phenomenon of Spatial Differentiation

Spatial differentiation refers to the visualization of price differentiation in different regions using maps based on the starting price, base price per kilometer, and return fare of taxis in each prefecture across the country. This includes the display of four basic fare differentiation phenomena: starting price, starting price per kilometer, base price per kilometer, and return empty kilometers. In addition to this, the final fare cost is calculated assuming a specific number of kilometers. Five kilometers is assumed to represent the distance traveled by residents nearby, 10 km represents the distance traveled by residents in medium proximity, 20 km represents the distance traveled by residents in long proximity, and 40 km represents the distance traveled by residents in long proximity generally to the airport.

The Law of Spatial Differentiation

The law of spatial differentiation refers to the characteristics of the phenomenon of spatial dispersion in taxi fares using specific methods. The methods used in this paper include a global differentiation index (GDI) taxi fare dispersion measures (Y. Wang et al., 2013), Global Moran’s I (Bivand et al., 2009), and kernel density analysis (Terrell & Scott, 1992). Taxi fares for prefecture-level cities can be either face data, using urban data to represent taxi fares for the whole city and its subordinate counties, or point data, representing only taxi fares in the urban center of the prefecture-level city.

The GDI of taxi fares provides a comprehensive indication of the degree of variation in the overall variation of taxi fares across prefectures nationwide. The main components are the coefficient of variation (CV), the Thiel index (T), the generalized entropy index (GE), and the Atkinson index (A) (Atkinson, 1970; Brown, 1998; Shorrocks, 1980; Wilson & Lockshin, 2003). The coefficient of variation measures the variability of a series of numbers independently of the unit of measurement used for these numbers (Abdi, 2010). Theil Index Decomposition is widely used in the testify and analysis of the total difference in one region and differences between regions (J. Sun et al., 2023). Generalized Entropy is a widely used index for studying income disparity, and the Atkinson index can be used to measure the degree of inequality in an economy (Biewen & Jenkins, 2006). Finally, the Gini coefficient (G) (Dorfman, 1979) is used to portray the degree of divergence. It is calculated as:

CV = \sqrt{\sum_{i = 1}^{n} {(x_{i} - μ)}^{2} / n} / n

T = \frac{1}{n} \sum_{i = 1}^{n} \frac{x_{i}}{μ} \log \frac{x_{i}}{μ}

(C = 0) : GE = \frac{1}{n} \sum_{i = 1}^{n} \log (\frac{μ}{x_{i}})

A = 1 - {[\frac{1}{n} \sum_{i = 1}^{n} {(\frac{x_{i}}{μ})}^{1 - ε}]}^{\frac{1}{1 - ε}}

G = \frac{S_{A}}{S_{A} + S_{B}}

Using information entropy to determine the weights as W_j. The GDI of taxi fares is then:

GDI = W_{1} * CV + W_{2} * T + W_{3} * GE + W_{4} * A

Where: x_i is the taxi fares of the ith prefecture-level city, μ is the average taxi fares, and n is the number of prefecture-level cities. ε is a parameter related to the sensitivity of the variance value and takes a range of values greater than 0. The larger the ε, the greater the weight given to the prefecture-level cities with relatively low taxi fares. ε often takes values of .5 and 2 (Atkinson, 1970), and the value of ε is taken as 2 for the calculation of this paper.

Spatial autocorrelation analysis of taxi fares can indicate spatial correlation and aggregation characteristics. The global Moran’s I (GMI) was used to measure the degree of spatial correlation of taxi fare indices at the prefecture level, calculated as:

GMI = \frac{\sum_{i = 1}^{n} \sum_{j = 1}^{n} W_{i} (x_{i} - \bar{x}) (x_{j} - \bar{x})}{S^{2} \sum_{i = 1}^{n} \sum_{j = 1}^{n} W_{ij}}

S^{2} = \sum_{i = 1}^{n} (x_{i} - \bar{x})^{2} / n

Z (I) = \frac{I - E (I)}{\sqrt{Var (I)}}

Where: n is the number of samples; x_i and x_j are the taxi fares of spatial units i and j respectively; W_ij is the spatial weight matrix of each prefecture, and the distance between prefectures is 1 within a set threshold distance, or 0 if greater than that distance. E(I) is the mathematical expectation; Var(I) is the number of variances. Generally when |Z| > 1.96, it means spatial autocorrelation exists at a 95% confidence level; |Z| > 2.58, it means significant at a 99% confidence level.

Kernel density analysis calculates the accumulation of the whole area using the input points, which allows a more accurate portrayal of the overall distribution characteristics of taxi fares. The calculation formula is:

f_{n} (x) = \frac{1}{n h_{n}} \sum_{i = 1}^{n} K (\frac{x - x_{i}}{h_{n}})

where: K( ) is the kernel function; n is the number of sample points in the bandwidth range; (xxi) denotes the distance from the estimated point to sample x_i; and he is the bandwidth and is a positive number related to n.

Taxi Fares Model

There are many factors affecting taxi fares, including transport costs, supply and demand, national policies, local development levels, etc. Among them, transport costs include vehicle acquisition costs, management service fees, vehicle maintenance fees, fuel costs, etc. Supply and demand mainly include the city population, the number of taxis in the city, the level of development of other public transport in the city, etc. National policies mainly include government-led pricing power, tax incentives, etc. The local development level mainly includes the city’s economic development level, per capita GDP, disposable income of residents, etc. Some of the above factors are difficult to quantify, and some of them vary from city to city. The GDP growth rate is an external factor influencing taxi travel. At the same time, the trend in the use of taxis as a means of transportation is closely related to the economic situation of individuals and the number of inhabitants (An et al., 2021; Kim et al., 2018). With the development of urbanization, the taxi market in large cities is rapidly expanding and developing (Qian & Ukkusuri, 2015). Therefore, a total of four factors, namely GDP of regional GDP, disposable income of urban residents, population density in urban areas, and urbanization level, were selected to establish a multiple linear regression model for taxi fares. The criterion for judging whether they are the main influencing factors is the significance level of each factor in this model. The expressions are:

y = a_{0} + a_{1} x_{1} + a_{2} x_{2} + \dots + a_{n} x_{n}

Where: x₁, x₂…, x_n are the factors affecting taxi fares; a₁, a₂,…, a_n are the regression coefficients of the factors affecting taxi fares; y is the taxi fare; and a₀ is the intercept.

Results

The Phenomenon of Spatial Differentiation in Taxi Fares

Taxi fares are composed of a starting price, a base price, and a return empty price.

Taxi fares in China range from RMB 4 to 14, with short-distance trips being more sensitive to the starting price. Low starting prices are mainly distributed in the central and western regions, while high starting prices are mainly found in the eastern coastal regions, especially in the Beijing-Tianjin, Yangtze River Delta, and Pearl River Delta metropolitan areas. In addition, some provincial capitals in central and western provinces have higher starting prices, such as Lanzhou, Urumqi, Lhasa, Xi’an, and Guiyang, which are significantly higher than other cities in the province. The number of kilometers included in the starting price varies, and the starting price per kilometer reflects whether the taxi is starting at a reasonable price. The starting price per kilometer for taxis in China ranges from RMB 1.67 to RMB 6. The starting price per kilometer is lower in the northeast and northwest regions, and higher in provinces such as Sichuan, Guangxi, Guangdong, and Zhejiang. The base price per kilometer is the price per kilometer of taxi travel beyond the starting kilometer, with medium and long-distance trips being more sensitive to the starting price. The distribution of the base price shows two clear areas of concentrated high values, in the Yangtze River Delta and Pearl River Delta regions, in addition to higher base prices in Beijing and Tianjin, and in Yunnan, a major tourist province, where some cities also have higher base prices. The divergence in base prices is mainly north-south, with the southern region being significantly higher than the northern region. Long-distance taxi rides beyond a certain distance are generally subject to an additional return fare or emptying charge. Distance regulations vary from province to province and municipality to municipality, and charges are generally increased by 50% on the base price, with a minimum increase of 20% and a maximum increase of 100%, with some municipalities also increasing the price in segments depending on the distance. Generally speaking, the further the return distance, the better the discount for the consumer. There is no clear pattern to the national taxi regulations on return air distance, with the closest being 3 km, with return air charges applied beyond the starting kilometer, and the furthest up to 25 km, mainly in Chongqing and some cities in Guangdong (Figure 2).

Figure 2.

The pattern of differential taxi fares.

The cost of a 5, 10, 20, and 40 km journey is calculated based on the starting price, base price, and return fare respectively. Five kilometers is the most frequent distance traveled by residents, and this distance is more related to the starting price, so the pattern of variation is roughly the same as the starting price. The average cost of a 5 km taxi journey in China ranges from RMB 7.4 to RMB 18.8, with an average cost of RMB 11.53, with high values in the Beijing-Tianjin region, the Yangtze River Delta and the Pearl River Delta, and a general trend of high in the southeast and low in the northwest, with some western provincial capitals spending more for 5 km. The cost of a 10, 20, and 40 km taxi ride is roughly the same, with the overall trend being higher in the south and lower in the north, with an average cost of RMB22.04, RMB47.54, and RMB99.80 respectively. The high-value areas are concentrated in Jiangsu, Zhejiang, and most of Guangdong, with some areas in Yunnan, Sichuan, and Hunan also higher, while in the north, apart from Beijing and Tianjin and individual cities in Shandong, Liaoning, and Jilin, most costs are lower than the national average (Figure 3).

Figure 3.

Differences of taxi fares in different distances.

Spatial Differentiation of Taxi Fares

Global Differentiation Index of Taxi Fares

Using the four elements of the GDI, the degree of divergence was calculated for taxis traveling 5, 10, 20, and 40 km nationwide. The results are shown in Table 1. Overall, the greater the distance traveled, the greater the overall measure of variation index and the greater the Gini coefficient. This indicates that for taxi fares across the country, the greater the distance traveled, the greater the variation in prices between cities.

Table 1.

GDI and Gini Coefficient.

Index	5 km	10 km	20 km	40 km
CV	.0062	.0114	.0241	.0526
T	.0102	.0069	.0052	.0049
GE	.0068	.0068	.0064	.0069
A	.0307	.0313	.0292	.0311
GDI	.0539	.0564	.0649	.0955
G	.3423	.4217	.4263	.4245

Spatial Autocorrelation Analysis

Taking 5 km as an example, the spatial correlation and clustering characteristics of taxi fares across the country were analyzed.

According to the calculated results, the Moran I result is 0.151521 with a z-score of 20.54, which means that the probability of randomly generating this clustering pattern is less than 1%, indicating that Chinese taxi fares in general show significant spatial clustering characteristics. Based on global spatial autocorrelation, local spatial autocorrelation can also be analyzed. Moran I has the problem of ignoring some of the potential spatial instability when assessing spatial autocorrelation. Local spatial autocorrelation analysis is required when further examining the local spatial clustering of high or low values of observations and which regional units contribute more to global autocorrelation. According to the results of local spatial autocorrelation, the H-H region of Chinese taxi fares is concentrated in the Beijing-Tianjin, Yangtze River Delta, and Pearl River Delta regions, while the L-L region is concentrated in the Northwest and Northeast regions, mainly including Gansu, Ningxia, Shaanxi, and parts of Inner Mongolia in the Northwest.

In addition, this paper calculates the local spatial correlation index Getis-OrdGi* and visualizes it spatially in ArcGIS software, using the natural fracture method to divide the national taxi fares into hot spot areas and cold spot areas. The hotspot areas of taxi fares in China are mainly located in the southeastern coastal areas, while the coldspot areas are mainly located in the northeast, north China, and northwest except Xinjiang, and the distribution of coldspot areas has an obvious divergence pattern from southeast to northwest (Figure 4).

Figure 4.

Local spatial autocorrelation aggregation and Getis-OrdGi*.

Kernel Density

Kernel density analysis was carried out of 5 and 40 km, and the results showed that: Firstly, the high kernel density areas are found in the Yangtze River Delta and the Pearl River Delta, but not in the Beijing-Tianjin region, indicating that taxi fares are lower around the Beijing-Tianjin region, while taxi fares are higher in the Yangtze River Delta and the Pearl River Delta as a whole. Secondly, the distribution of the kernel density of taxi fares shows a decreasing trend from the southeast coast to the northwest inland. Finally, the distribution of kernel density tends to be spread out with the provincial capitals as the high-value centers (Figure 5).

Figure 5.

Kernel density of 5 and 40 km.

Multiple Linear Regression Model for Taxi Fares

Taxi fares are determined by a combination of factors, with the standard of living of the local population being the most important factor in determining the price of a taxi in a given area, with urban disposable income per capita being the most representative of the standard of living. In addition to this, regional GDP, urbanization levels, and urban population density are also included as factors affecting taxi fares. Accordingly, a scatter diagram (Figure 6) was generated for the relationship between taxi fares and urban disposable income per capita, regional GDP, urbanization level, and urban population density. The graph shows that the above four factors are all positively correlated with taxi fares and that the individual factors are more representative of the differences in taxi fares, thus verifying the reasonableness of the selection of these four influencing factors. Among them, the R² of urban disposable income per capita is the largest, which proves that this factor has the greatest influence on taxi fares.

Figure 6.

Scatter diagram of taxi fares and influencing factors.

The different factors interact with each other and jointly influence taxi fares. Based on the above four factors, a multivariate linear regression model was established to determine the degree of influence of different factors on taxi fares and to provide model support for the adjustment of taxi fares in different regions under different factors.

The model regression yielded a goodness of fit R of .73 and an adjusted R² of .528. The F value was 96.354, with a significance level of .000, at a confidence level of α = .01, F is much greater than the corresponding F distribution value, indicating that the regression model is highly significant. It is not difficult to find from the table that the collinearity diagnosis results (VIF values) of each explanatory variable are less than 5, indicating that there is no collinearity relationship among the variables. The sig. value of each variable was less than .05, indicating that each variable had a significant influence on the explanatory model (X. Zhang et al., 2023) Also, the absolute values of the explanatory model T-values were greater than 1.96, further indicating that each explanatory variable had a significant impact on the explanatory model. Finally, the sig. values of the explanatory models were all greater than .05, indicating that the explanatory models were statistically significant (Table 2).

Table 2.

Multiple Linear Regression Analysis of Taxi Fares.

Model	Classification of indicators	Variable	B	T	Sig.	VIF
		(Constant)	6.065	14.52	0
Dependent variable: taxi fare of 5 km	GDP	Total GDP of the city (X1)	.007	2.294	.022	2.465
	Urbanization	Urbanization level of the city (X2)	.032	4.665	0	1.802
	Income	Disposable income of urban residents of the city (X3)	.862	5.895	0	2.504
	Population density	Population density in urban areas of the city (X4)	.124	4.052	0	1.568

A robustness test is necessary. (1) Excluding some city samples. The cities in this paper’s sample include both ordinary prefecture-level cities, as well as municipalities and provincial capitals. Since municipalities and provincial capitals have significantly higher population mobility than ordinary prefecture-level cities, the resulting high demand for rides may affect taxi fares. In this paper, all the samples of municipalities and provincial capitals are deleted, and only the samples of ordinary prefecture-level cities are retained for regression. The results show that the significance levels of the factors remain high, and thus the results are robust. (2) Variable substitution. The dependent variable in this paper is the taxi fare, which was chosen to be the cost of traveling 5 km when the regression was conducted. To further test the robustness of the results, the dependent variable is replaced with the cost required for a taxi to travel 10, 20, and 40 km. After re-regression based on the dependent variable, it is found that the significance of the explanatory variables remains unchanged, proving the robustness of the model. (3) Test method replacement. This paper defines a dummy variable for whether taxi fares are higher than average and uses it as an explanatory variable in a Logit regression, and the main conclusions remain the same.

Endogeneity issue and instrumental variable. There may be an endogeneity problem in this paper. The main reason is that there is no control for any alternative mode of transportation that may lead to missing variables. Taxi fare, as income of services, is contributing to local GDP and resident income. There is a potential reverse causality between the explanatory variables and the explained variables. To mitigate estimation bias due to endogeneity, instrumental variable regressions are conducted in this paper. Referring to related studies, the water quality of the city is selected as an instrumental variable for city GDP (Russ et al., 2022), and the air quality of the city is selected as an instrumental variable for the disposable income of urban residents (L. Wang et al., 2018). The regression results in this paper remain robust after accounting for model endogeneity issues.

A taxi fares model of China was developed as follows:

y = 6.065 + 0.007 x_{1} + 0.032 x_{2} + 0.862 x_{3} + 0.124 x_{4}

Where y is the taxi fare, x₁ is the regional GDP, x₂ is the level of urbanization, x₃ is the per capita disposable income of urban residents, and x₄ is the population density of urban areas. For every 10 billion yuan increase in regional GDP, the taxi fare increases by RMB 0.007, for every 1% increase in urbanization level, the taxi fare increases by RMB 0.032, for a 10,000 yuan increase in urban residents’ per capita disposable income, the taxi fare increases by RMB 0.862, and for a thousand people per square kilometer increase in urban population density, the taxi fare increases by RMB 0.124.

The constant, or intercept, is 6.065, which roughly indicates that in even the least economically developed areas of China, a 5 km taxi ride fare is more than RMB 6. The biggest influence on taxi fares is the per capita disposable income of urban residents, which is the factor most closely related to residents’ lives. In areas where the per capita disposable income is higher, taxi fares are already higher and residents are less sensitive to the strength of taxi fare increases when they are made.

Discussion

Taxi fares are a matter of people’s lives and have an important impact on social stability. On the one hand, residents want cheaper taxi fares so that they can reduce their travel costs. On the other hand, taxi drivers want higher taxi fares so that they can increase their income. From the above analysis, it can be found that the overall pattern of taxi fares in China matches the level of economic development. Taxi fares are more expensive in medium and large cities and cheaper in smaller cities. Taxi fares are more expensive in the economically developed regions of Beijing, Tianjin, and Hebei, the Yangtze River Delta, and the Pearl River Delta, and cheaper in the northeast and western regions.

The biggest influence on taxi fares is the per capita disposable income of urban residents, the amount of income directly determines the level of consumption of residents. The other three items have a smaller impact on taxi fares compared to the per capita disposable income of urban residents. Relevant studies have shown that regions with greater per capita disposable income and per capita consumption expenditure have more taxi orders and tend to have higher taxi fares (B. Hu et al., 2021; H. Yang et al., 2000). Regional GDP is a reflection of the level of comprehensive economic development of a region. However, there are also resource-based cities such as Tangshan, Daqing, and Jiaozuo where GDP is high but taxi fares are low. In cities with higher GDP, travel costs are higher and taxi fares are more expensive (Salanova et al., 2011). Tourism cities such as Sanya, Haikou, and Lijiang where GDP is high but taxis mainly serve foreign tourists and thus are higher. Rapid economic growth and urbanization have dramatically changed China’s urban public transportation, fuel prices, parking fees, and taxi fares (C. Sun et al., 2014). The level of urbanization has a smaller impact on taxi fares for the same reasons as regional GDP, with some small and medium-sized cities such as Karamay, Wuhai, and Jiayuguan having a higher level of urbanization but lower taxi fares, and less urbanized ones such as tourist cities. The relation of taxi cost versus bus cost grows with the density of the city, due to the economies of scale of the Mass Public Transport (Salanova et al., 2011). Urban population density reflects the concentration of population, the greater the population density in urban areas the higher the demand for taxis, if there is a long-term supply shortage, the phenomenon of raising taxi fares may occur, Henan, Shandong, and other places in some urban areas with high population density, but the labor market is sufficient, taxi fares are also lower, tourist cities with a small resident population, low population density, but will taxi fares on the high side of the phenomenon.

Based on analyzing the spatial differentiation of taxi fares in China, this study explores what factors affect taxi fares. Generally speaking, the spatial transmission mechanism of urban taxi fares is more significant at the micro level; however, a macro-level study helps to grasp the pattern of spatial differentiation in general. The main contribution of the study is to make up for the lack of analyzing the spatial differentiation of taxi fares as a research object from the macro level. In the future, with the adjustment of taxi fares in different cities, related studies need to be followed up and the results of the macro and micro studies can be compared. In conclusion, taxi fares in a city are determined by a variety of factors, and the model built in this paper only takes into account some of these factors, and the taxi price model thus built is only a correlation rather than a causal relationship. Based on the four factors selected in this paper, a judgment can be made on taxi fares, and the data model can be used to support the rationality of taxi fare adjustments. The divergent pattern of taxi fares and the setting of price increases in a particular city are influenced by multiple factors, of which only four are discussed in this paper. In addition, macro policy regulation, vehicle transportation costs, and supply and demand games can all affect taxi fares, and further exploration is needed in the future where this paper has not focused.

Conclusions

Based on the prefecture-level city scale, the distribution pattern of taxi fares in various cities across the country was analyzed and some indicators were selected to explore the factors affecting taxi fares, with the following conclusions:

The starting price of taxis across the country ranges from RMB 4 to RMB 14, and the price per kilometer ranges from RMB 1 to RMB 3. In addition to this, there are also return emptying fees, night-time driving fees, and waiting fees. The distribution pattern mainly includes taxi fares being higher in provincial capitals than in non-capital cities, and higher in southern cities than in northern cities, with high values concentrated in the Beijing-Tianjin metropolitan area, the Yangtze River Delta metropolitan area, and the Pearl River Delta metropolitan area.

The greater the distance traveled by taxi fares nationwide, the greater the overall divergence measure index and the greater the Gini coefficient. National taxi fares in general show significant spatial agglomeration characteristics and form two low-low agglomeration and three high-high agglomeration centers. The high nuclear density areas are found in the Yangtze River Delta and the Pearl River Delta, with the provincial capitals as the high-value centers of dispersion.

The pattern of taxi price divergence across the country is related to several factors, the most influential of which is the per capita disposable income of urban residents, in addition to regional GDP, urbanization rate, and urban population density, which also have a greater impact on taxi fares.

In conclusion, there is a clear spatial differentiation of taxi fares in China, and this differentiation becomes larger as the distance traveled by taxis increases. Taxi fares in China are most affected by disposable income per urban resident.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest concerning the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research was funded by the National Natural Science Foundation of China, Grant no. 42371264.

ORCID iD

Dou Wenkang

Data Availability Statement

The data that support the findings of this study are openly available in provincial and municipal statistical yearbooks and bulletins.

References

Abdi

(2010). Coefficient of variation. Encyclopedia of Research Design, 1(5), 169–171.

Wang

(2021). A system dynamics model for urban taxi price simulation. Transport, 36(5), 395–405.

Atkinson

A. B.

(1970). On the measurement of inequality. Journal of Economic Theory, 2(3), 244–263.

Biewen

Jenkins

S. P.

(2006). Variance estimation for generalized entropy and Atkinson inequality indices: The complex survey data case. Oxford Bulletin of Economics and Statistics, 68(3), 371–383.

Bivand

Müller

W. G.

Reder

(2009). Power calculations for global and local Moran’s I. Computational Statistics & Data Analysis, 53(8), 2859–2872.

Brown

C. E.

(1998). Coefficient of variation applied multivariate statistics in geohydrology and related sciences (pp. 155–157). Springer.

Cetin

Deakin

(2019). Regulation of taxis and the rise of ridesharing. Transport Policy, 76, 149–158.

Chan

(2001). Location theory and decision analysis. Springer.

Chen

Zhang

Yao

Hong

Guan

(2019). Exploring the spatial differentiation of urbanization on two sides of the Hu Huanyong Line—Based on nighttime light data and cellular automata. Applied Geography, 112, 102081.

10.

Choi

Kim

Kwak

Park

Lee

(2023). Measuring taxi ridesharing effects and its spatiotemporal pattern in Seoul, Korea. Travel Behaviour and Society, 30, 148–162.

11.

Deng

(2011, June 24–26). Spatiotemporal structure of taxi services in Shanghai: Using exploratory spatial data analysis [Conference session]. 2011 19th International Conference on Geoinformatics, Shanghai, China. IEEE.

12.

Dorfman

(1979). A formula for the Gini coefficient. The Review of Economics and Statistics, 61(1), 146–149.

13.

Feng

Somenahalli

Bian

Zhou

Wang

(2023). A study of spatio-temporal differentiation characteristics and driving factors of Shaanxi Province’s traditional heritage villages. Sustainability, 15(10), 7797.

14.

Gallo

(2018). Improving equity of urban transit systems with the adoption of origin-destination-based taxi fares. Socio-Economic Planning Sciences, 64, 38–55.

15.

Gan

Wang

Sun

Shi

(2013, August 3–9). Optimal pricing for improving the efficiency of taxi systems [Conference session]. Proceedings of the 23rd International Joint Conference on Artificial Intelligence (IJCAI’13), Beijing, China.

16.

Gómez-Lobo

Tirachini

Gutierrez

(2022). Optimal prices for ride-sourcing in the presence of taxis, public transport, and car competition. Transportation Research Part C: Emerging Technologies, 137, 103591.

17.

Grau

J. M. S.

Estrada

(2019). Social optimal shifts and fares for the Barcelona taxi sector. Transport Policy, 76, 111–122.

18.

Guo

Sun

Schonfeld

(2021). Time-dependent transit fare optimization with elastic and spatially distributed demand. Transportation Research Part A: Policy and Practice, 148, 353–378.

19.

Wang

Lin

Tang

(2018). Pricing and penalty/compensation strategies of a taxi-hailing platform. Transportation Research Part C: Emerging Technologies, 86, 263–279.

20.

Hoover

(2022). Nudges as norms: Evidence from the NYC taxi cab industry. Journal of Economic Psychology, 92, 102535.

21.

Hou

Deng

Yao

(2022). Coordinated relationship between urbanization and grain production in China: Degree measurement, spatial differentiation, and its factors detection. Journal of Cleaner Production, 331, 129957.

22.

Zhang

Ding

Zhang

Dong

Sun

(2021). Research on the coupling degree of regional taxi demand and social development from the perspective of job–housing travels. Physica A: Statistical Mechanics and Its Applications, 564, 125493.

23.

Thill

J.-C.

(2019). Predicting the upcoming services of vacant taxis near fixed locations using taxi trajectories. ISPRS International Journal of Geo-Information, 8(7), 295.

24.

Huang

Liu

Wang

(2021). Measuring accessibility based on improved impedance and attractive functions using taxi trajectory data. Sustainability, 13(1), 112.

25.

Jiang

(2015). Geospatial analysis requires a different way of thinking: The problem of spatial heterogeneity. GeoJournal, 80, 1–13.

26.

Jin

Shi

Guo

Shi

(2020). Designing a spatial pattern to rebalance the orientation of development and protection in Wuhan. Journal of Geographical Sciences, 30(4), 569–582.

27.

Kamga

Yazici

M. A.

Singhal

(2015). Analysis of taxi demand and supply in New York City: Implications of recent taxi regulations. Transportation Planning and Technology, 38(6), 601–625.

28.

Kang

Duan

Yan

(2022). Spatial differentiation and impact factors of tourism development: A case study of the central plains, China. Sustainability, 14(12), 7313.

29.

Kim

Baek

Lee

J.-D.

(2018). The creative destruction of the sharing economy in action: The case of Uber. Transportation Research Part A: Policy and Practice, 110, 118–127.

30.

Szeto

Zou

(2022). Optimal fare and fleet size regulation in a taxi/ride-sourcing market with congestion effects, emission externalities, and gasoline/electric vehicles. Transportation Research Part A: Policy and Practice, 157, 215–243.

31.

Dong

Shen

Lang

(2017). Examining the interaction of taxi and subway ridership for sustainable urbanization. Sustainability, 9(2), 242.

32.

Z.-T.

Nie

W.-P.

Cai

S.-M.

Zhao

Z.-D.

Zhou

(2023). Exploring the topological characteristics of urban trip networks based on taxi trajectory data. Physica A: Statistical Mechanics and Its Applications, 609, 128391.

33.

Liao

Zuo

Law

Wang

Zhang

(2022). Spatial differentiation, influencing factors, and development paths of rural tourism resources in Guangdong Province. Land, 11(11), 2046.

34.

Liu

Han

Chen

X. M.

Ong

G. P.

(2019). Spatial-temporal inference of urban traffic emissions based on taxi trajectories and multi-source urban data. Transportation Research Part C: Emerging Technologies, 106, 145–165.

35.

Liu

Vergara-Cobos

Zhou

(2019). Pricing schemes and seller fraud: Evidence from New York City taxi rides. The Journal of Industrial Economics, 67(1), 56–90.

36.

Liu

Fang

(2022). Spatial pattern change and influencing factors of industrial eco-efficiency of Yangtze River Economic Belt (YREB). SAGE Open, 12(3), 21582440221113849.

37.

Luo

Dong

Dou

Zhang

Ren

Sun

Yao

(2017). Analysis on spatial-temporal features of taxis’ emissions from big data-informed travel patterns: A case of Shanghai, China. Journal of Cleaner Production, 142, 926–935.

38.

Lyu

Wang

Feng

(2023). A multiscale spatial analysis of taxi ridership. Journal of Transport Geography, 113, 103718.

39.

Mao

Liu

(2016). Mining spatiotemporal patterns of urban dwellers from taxi trajectory data. Frontiers of Earth Science, 10, 205–221.

40.

Nie

W.-P.

Cai

S.-M.

Zhao

Z.-D.

Gao

F.-J.

Z.-T.

Zhou

(2023). Examining the impact of urban-rural spatial structure on mobility networks: A case study of taxis. Cities, 141, 104512.

41.

North

D. C.

(1955). Location theory and regional economic growth. Journal of Political Economy, 63(3), 243–258.

42.

Phiboonbanakit

Horanont

(2016, March 20–22). Who will benefit from the new taxi fare rate? Discerning the real driving from taxi GPS data [Conference session]. 2016 7th International Conference of Information and Communication Technology for Embedded Systems (IC-ICTES), Bangkok, Thailand. IEEE.

43.

Qian

Ukkusuri

S. V.

(2015). Spatial variation of the urban taxi ridership using GPS data. Applied Geography, 59, 31–42.

44.

Rahimi

Sadeghi-Niaraki

Ghodousi

Choi

S.-M.

(2021). Modeling population spatial-temporal distribution using taxi origin and destination data. Sustainability, 13(7), 3727.

45.

Russ

Zaveri

Desbureaux

Damania

Rodella

A.-S.

(2022). The impact of water quality on GDP growth: Evidence from around the world. Water Security, 17, 100130.

46.

Salanova

J. M.

Estrada

Aifadopoulou

Mitsakis

(2011). A review of the modeling of taxi services. Procedia-Social and Behavioral Sciences, 20, 150–161.

47.

Schaller

(1998). Issues in fare policy: Case of the New York taxi industry. Transportation Research Record, 1618(1), 139–142.

48.

Shorrocks

A. F.

(1980). The class of additively decomposable inequality measures. Econometrica: Journal of the Econometric Society, 46(3), 613–625.

49.

Silveira-Santos

Papanikolaou

Rangel

Manuel Vassallo

(2023). Understanding and predicting ride-hailing fares in Madrid: A combination of supervised and unsupervised techniques. Applied Sciences, 13(8), 5147.

50.

Siyal

A. W.

Hongzhuan

Gang

(2021). From consumer satisfaction to recommendation of mobile app-based services: An overview of mobile taxi booking apps. SAGE Open, 11(1), 1–14.

51.

Sun

Zheng

Wang

(2014). Restricting driving for better traffic and clearer skies: Did it work in Beijing? Transport Policy, 32, 34–41.

52.

Sun

Shi

(2023). A research of the evaluation of preschool education resource allocation level and spatio-temporal differences: Based on repeated indicators method and Theil index. Heliyon, 9(6), 1–15.

53.

Tang

Sun

Kan

Ren

Cheng

(2017). Uncovering distribution patterns of high-performance taxis from big trace data. ISPRS International Journal of Geo-Information, 6(5), 134.

54.

Terrell

G. R.

Scott

D. W.

(1992). Variable kernel density estimation. The Annals of Statistics, 20, 1236–1265.

55.

Thathsarani

Jianguo

Alariqi

(2023). How do demand, supply, and institutional factors influence SME financial inclusion: A developing country perspective. SAGE Open, 13(3), 1–13.

56.

Ulahannan

Birrell

(2022). Designing better public transport: Understanding mode choice preferences following the COVID-19 pandemic. Sustainability, 14(10), 5952.

57.

Wang

Fan

Chai

Wang

Long

Gao

(2022). Spatial-temporal evolution and influencing mechanism of traffic dominance in Qinghai-Tibet Plateau. Sustainability, 14(17), 11031.

58.

Wang

Huang

Zeng

(2019). Revealing spatial-temporal characteristics and patterns of urban travel: A large-scale analysis and visualization study with taxi GPS data. ISPRS International Journal of Geo-Information, 8(6), 257.

59.

Wang

Huang

Jing

(2020). Competition between high-speed trains and air travel in China: From a spatial to spatiotemporal perspective. Transportation Research Part A: Policy and Practice, 133, 62–78.

60.

Wang

Sun

(2013, January 13–17). Game theory-based optimization of taxi fare in the taxi market of China [Paper presentation]. Transportation Research Board 92nd Annual Meeting, Washington, DC, USA.

61.

Wang

Fang

Law

(2018). Effect of air quality in the place of origin on outbound tourism demand: Disposable income as a moderator. Tourism Management, 68, 152–161.

62.

Wang

Zhang

Zhou

Sun

(2022). Spatial-temporal changes in ecosystem services and social-ecological drivers in a typical coastal tourism city: A case study of Sanya, China. Ecological Indicators, 145, 109607.

63.

Wang

Lin

Zhang

(2016). Spatial differentiation patterns and influencing mechanism of housing prices in China: Based on data of 2872 counties. Acta Geographica Sinica, 71, 1329–1342. (in Chinese)

64.

Wang

Yang

Gao

H. O.

(2016). Pricing strategies for a taxi-hailing platform. Transportation Research Part E: Logistics and Transportation Review, 93, 212–231.

65.

Wang

Fang

Sheng

(2013). Spatial differentiation and model evolution of housing prices in Yangzhou. Acta Geographica Sinica, 68, 1082–1096. (in Chinese)

66.

Wilson

Lockshin

(2003). Using the Thiel index coefficient to analyze variation in wine consumption habits. University of South Australia.

67.

Wong

Szeto

(2022). The effects of peak hour and congested area taxi surcharges on customers’ travel decisions: Empirical evidence and policy implications. Transport Policy, 121, 78–89.

68.

Wong

Yang

Szeto

Wong

(2020). The effects of accessible taxi service and taxi fare subsidy scheme on the elderly’s willingness to travel. Transport Policy, 97, 129–136.

69.

Zhang

Tian

Wang

Hua

(2020). Spatial differentiation and network externality in pricing mechanism of online car-hailing platform. International Journal of Production Economics, 219, 275–283.

70.

Wyman

K. M.

(2017). Taxi regulation in the age of Uber. The N.Y.U. Journal of Legislation and Public Policy, 20, 1.

71.

Xiao

Ullah

Zhang

(2023). Poverty alleviation efficiency of tourism and its spatiotemporal differentiation in Jiangxi Province of China based on the DEA model. SAGE Open, 13(2), 1–13.

72.

Xing

(2023). Spatial-temporal differentiation and spatial interaction effect analysis of residents’ consumption capacity and consumption upgrading in China. Systems, 11(5), 219.

73.

Wong

Yang

Tong

C.-O.

(1999). Modeling level of urban taxi services using neural network. Journal of Transportation Engineering, 125(3), 216–223.

74.

Yang

Fung

Wong

K. I.

Wong

S. C.

(2010). Nonlinear pricing of taxi services. Transportation Research Part A: Policy and Practice, 44(5), 337–348.

75.

Yang

Lau

Y. W.

Wong

S. C.

H. K.

(2000). A macroscopic taxi model for passenger demand, taxi utilization and level of services. Transportation, 27, 317–340.

76.

Yang

Wong

S. C.

Wong

K. I.

(2002). Demand-supply equilibrium of taxi services in a network under competition and regulation. Transportation Research Part B: Methodological, 36(9), 799–819.

77.

Yang

Wong

S. C.

Sze

N. N.

(2014). Returns to scale in the production of taxi services: An empirical analysis. Transportmetrica A: Transport Science, 10(9), 775–790.

78.

Yang

Bao

Wang

Liu

(2021). Measurement of urban-rural integration level and its spatial differentiation in China in the new century. Habitat International, 117, 102420.

79.

Yang

Song

Wang

(2018). Investigation on structural and spatial characteristics of taxi trip trajectory network in Xi’an, China. Physica A: Statistical Mechanics and Its Applications, 506, 755–766.

80.

Zhuo

Zhang

(2023). Research on the spatial correlation and influence factors of regional internet finance in China. SAGE Open, 13(3), 1–11.

81.

Zhang

Shi

Zhu

Wen

(2022). Influence of POI accessibility on temporal–spatial differentiation of housing prices: A case study of Hangzhou, China. Journal of Urban Planning and Development, 148(4), 04022037.

82.

Zhang

Ukkusuri

S. V.

Yang

(2018). Modeling the taxi drivers’ customer-searching behaviors outside downtown areas. Sustainability, 10(9), 3003.

83.

Zhang

Ukkusuri

S. V.

(2020). Understanding spatiotemporal variations of ridership by multiple taxi services. ISPRS International Journal of Geo-Information, 9(12), 757.

84.

Zhang

Han

Tang

Chen

(2023). Spatial distribution characteristics and driving factors of tourism resources in China. Land, 12(5), 1029.

85.

Zhao

Liu

(2018). Spatial differentiation and influencing mechanism of medical care accessibility in Beijing: A migrant equality perspective. Chinese Geographical Science, 28, 353–362.

86.

Zhou

Rong

Yang

Zhang

Khezerlou

A. V.

Zheng

Shafiq

Liu

A. X.

(2018). Optimizing taxi driver profit efficiency: A spatial network-based Markov decision process approach. IEEE Transactions on Big Data, 6(1), 145–158.

Spatial Pattern and Driving Mechanism of Urban Taxi Fares in China

Abstract

Keywords

Introduction

Materials and Methods

Study Area and Data Sources

Methods and Formulas

The Phenomenon of Spatial Differentiation

The Law of Spatial Differentiation

Taxi Fares Model

Results

The Phenomenon of Spatial Differentiation in Taxi Fares

Spatial Differentiation of Taxi Fares

Global Differentiation Index of Taxi Fares

Spatial Autocorrelation Analysis

Kernel Density

Multiple Linear Regression Model for Taxi Fares

Discussion

Conclusions

Footnotes

Declaration of Conflicting Interests

Funding

ORCID iD

Data Availability Statement

References