On a quiet night in May 2009, in the beautiful coastal city Rivertown in a wealthy eastern province, home to less than a million people, a group of men was on the move. They knocked on the door of each name on a list of more than twenty “uncooperative targets.” They sat the target down, talked him through how important his cooperation would be for the wellbeing of everyone, and casually mentioned the occupation, work locations, and schools of the target’s spouse, siblings, and children. They then listed some problems the target was having: debts, tax evasion, traffic law violations, and sometimes lawsuits. They kindly expressed regrets about how inconvenient it was for everyone to have to meet up at such a late hour, and how much they wished they did not need to show up at the target’s house. Eventually they pulled out a contract, and nicely suggested the target should think things over one more time, and not let personal interests trump the greater good. They then left the target’s home and moved on to the next name on the list.

In July, two months after this incident, all private bus firms and individual bus owners in Rivertown agreed to sell their bus firms to the city’s sole local state-owned bus firm. The group of men was a taskforce set up by the city government. It comprised officials from the city’s Traffic Bureau, Police Bureau, Department of Urban Management, and the state-owned bus firm. The targets they visited were owners of private bus firms in the city. It had all begun half a year before, in 2008, when Rivertown’s government decided to deprivatize the urban bus sector in preparation for a “public transportation system upgrade.” The upgrade plan included introducing new bus models, building the first light rail system in the city, and re-routing and cutting the existing bus lines around this light rail system. None of these changes was negotiated with the private bus firms or specified in their contracts with the city government, and the firms resisted strongly. Most of the private bus owners still had more than eight years left on their contract. Per the government’s request, a lot of them had just replaced many old buses with new, air-conditioned buses not long before. They also found the re-routing without negotiation or fair compensation unacceptable. In response, they organized strikes and reported the city government to the provincial Office of Letters and Visits.Footnote ¹ In March 2009, the city government’s patience ran thin. With an executive order to “get things done quickly,” the taskforce was set up and sent into action. After twenty-five years of private provision of urban bus services in Rivertown, deprivatization of the entire sector was completed within two months.

When two officials from that taskforce described this process of “persuasion” to me, they meant to emphasize how difficult their job was. They did not like being on the deprivatization taskforce. They had to work at night and risked revenge from the private bus firm owners who were usually well connected in the city – two of the bus firm owners also ran an underground casino business and had “muscles.” They had no choice but to adopt the persuasion style of “those land expropriation and house demolition jobs,” because private entrepreneurs “just don’t care about the big picture” – referring to the city’s big plans in the public transportation sector.Footnote ²

And many more private bus owners all over China did not get the big picture. In all the fifteen cities where I conducted interviews, all of them deprivatized the urban bus sector, most using some form of coercion. Nationwide, deprivatization of the bus sector became a trend from the mid 2000s, going against the national government’s call for more private participation in the sector. This is because the urban bus sector, a mundane foot soldier of urban life, had quietly become a top choice for visibility projects and selectoral campaigns.

A Brief History of Marketization of China’s Urban Buses

Private firms once dominated the urban bus sector in Chinese cities. This was among the first public service sectors opened to private investors in China. As the economy grew rapidly in the 1990s, demand for public transportation in cities quickly increased. However, city governments did not have the fiscal capacity to meet the increasing demand, especially after the tax reform in 1994.Footnote ³ In 1995, joining the global new public management movement of deregulation and privatization of public sectors, the Chinese national government allowed marketization in various urban public service sectors, including urban buses. Marketization of urban buses mostly took the form of government franchising, with the government allowing private firms the right to use a public domain (roads) and sell public services (within-city bus transit) to the public.Footnote ⁴ A standard franchise contract between local city governments and private bus firms had a term of twenty years. Under the contract, private bus firms had to obtain approval from the city’s Traffic Bureau before opening new bus routes. They also did not have control over bus fares, which were solely decided by the city’s Commodity Price Bureau. Private bus firms did not share their income with the city government. City governments were not shareholders in these private bus firms. Instead, the firms paid a fixed fee for the “special franchise right” when signing the contract and yearly taxes on profits.

Quickly, Chinese cities invited domestic and foreign private bus firms into the sector to start new bus firms. My data collection from 288 cities shows 233 cities (81 percent) privatized their urban bus sectors at some point between 1996 and 2016 (Figure 4.1). On average, each city had eight private bus firms and one public bus firm after privatization. Private urban buses became an important component of China’s thriving urban life.

Figure 4.1 Ownership change in China’s urban bus sector: 288 cities, 1996–2016

Things took a turn in the early 2000s. Going against the national government’s continued call for private actors in municipal public service sectors, most Chinese cities started to deprivatize the once dominant private bus firms in the mid 2000s. As shown in Figure 4.1, only 19 percent of cities still had the urban bus sector under private control by the end of 2016, compared with 81 percent in 2006, and 27 percent in 1996. There is also local variation between cities in how long the bus sector was opened to private firms, ranging from a mere two years to twenty years, with an average lifespan of 11.5 years of private ownership.

A Puzzling Sectoral Choice for Deprivatization

The urban bus sector is not a typical choice for systematic state takeover. The phenomenon of deprivatization is still understudied, but scholars have identified sectors vulnerable to state takeover, including utilities, oil, mining, rail, telecommunications, banking, and sometimes media. For monopolistic public service sectors such as water supply, private ownership might provide less efficient and more expensive public goods than under public ownership, prompting the state to take over these sectors to provide cheaper services (Ohemeng and Grant Reference Ohemeng and Grant2008). Sectors that are politically sensitive to the presence of foreign firms, such as banking, rail, telecommunications and natural resources, are likely to be nationalized to avoid foreign dependency (Jodice Reference Jodice1980; Shafer Reference Shafer1983; Kobrin Reference Kobrin1984). Sometimes sectors such as oil and mining are expropriated simply because they possess profitable assets that can produce long term returns for the state (Guriev, Kolotilin, and Sonin Reference Guriev, Kolotilin and Sonin2011; Hajzler Reference Hajzler2012). An authoritarian state might also deprivatize politically important sectors, such as media, to control the public (Gehlbach and Sonin Reference Gehlbach and Sonin2014). However, the urban bus sector does not fit into any of these sectoral profiles. It is not monopolistic in nature and in practice; it is not a politically sensitive sector crucial to national security; its assets – the buses – have high discount rates and usually lose most of their value within six years; and it does not produce high profit margins owing to heavily controlled bus fares. For these reasons, most countries that have privatized their bus services still have this sector in private hands (Finn and Mulley Reference Finn and Mulley2011). Even in the case of China, the trend of deprivatization of urban buses is going against the national government’s call for more private participation in the sector.

Why do Chinese local governments deprivatize urban buses, a non-monopolistic, thin-profit, and nonstrategic public service sector? By examining local variations and timing variations of deprivatization of the urban bus sector across Chinese cities, this chapter provides a new explanation for deprivatization in the authoritarian context. The end of marketization in the urban bus sector began when the sector became a political playground for visibility projects, part of selectoral campaigns. Deprivatization of the urban bus sector became primarily a political measure whereby local officials removed private bus firms that resisted visibility projects, rather than a policy borne out of economic or social concerns.

Visibility Projects and the Increasingly Inconvenient Private Firms

The urban bus sector became a popular choice for visibility projects in the mid 2000s along with several other public service sectors, after an earlier generation of visibility projects such as overly grand government buildings and excessive airports fell out of favor with central leaders in 2004.Footnote ⁵ This change, along with the center’s policy shift to sustainable development instead of economic growth alone, pushed local government officials to look for new visibility projects in line with the center’s new preference. The urban bus sector was chosen against this background.

The urban bus sector was chosen primarily because it is a highly visible sector. Chinese city governments call urban buses “the moving window” of a city. The sector’s key facilities – buses and bus stops – are highly exposed, and every change in these facilities is visible, making them a great vehicle for city leaders to showcase their ability in city management. A Bureau of Urban Management official said jealously in an interview that “urban buses are so easily used to show achievement. You change the buses, you add buses, you build lanes, anything you do in this sector makes the city immediately look different. What other sectors can compare to this?”Footnote ⁶ Furthermore, buses provide a quasi-public service. As such, local officials can frame projects in this sector as beneficial to the general public, rather than to a select group of beneficiaries. This gives these projects more of an appearance of being genuine efforts to improve social welfare, however ostentatious or unsustainable they might later prove to be.

Justification for launching visibility projects in the urban bus sector became available when the national government launched a series of central environmental initiatives from 2002. As environmental degradation worsened in China, the national government issued four “guidance and suggestions” documents (zhidao yijian) in 2002, 2005, 2012, and 2015, calling for further development of the public transportation sector to replace private vehicles so as to reduce exhaust fumes and to alleviate traffic congestion. These central initiatives are not incorporated into the formal cadre evaluation, they do not provide financial support to local governments, and they are open to local interpretation. Basically, these central suggestions are vague and optional, rather than specific and mandatory. Visibility projects in this sector thus present little political risk of inadvertently defying or challenging the authority of higher-level officials.

Therefore, local officials turned their eyes to the urban bus sector following these initiatives. In interviews with government officials and bus firm managers in fifteen cities, when asked about what prompted sectoral upgrades and changes in the urban bus sector, all interviewees mentioned it was in response to these central initiatives. Only a few officials also mentioned local development needs. With justification provided by the national government, and as a highly visible sector, the urban bus sector became a popular choice for visibility projects.

There are three major visibility projects in the urban bus sector, as confirmed by interviews with both government officials and bus firms: (1) bus beautification projects; (2) bus upgrade projects; and (3) construction of Bus Rapid Transit (BRT) systems.

First are bus beautification projects, often associated with a city’s effort to apply for national city title awards that have a long tradition under CCP rule. Interviewees identified two important national titles that require cooperation from bus operators to beautify their buses: the “National Civilized City” (quan guo wenming chengshi) and “National Public Transportation City” (gongjiao dushi).Footnote ⁷ The central government awards these titles to multiple cities at a time based on whether the city is orderly, clean, beautiful, and civilized. Winning a title helps city leaders stand out in urban management, and government interviewees consistently mentioned that winning a title is “an important line” on city leaders’ resumes and work reports, and leaders “care about these titles very much.”Footnote ⁸

Applications for these city titles focus on the looks of the buses, and in the run-up, city officials often embark on a project of cleaning and repainting dirty buses, taking advertisements off them, and replacing old-looking buses. Bus beautification projects do not have any obvious social welfare implications in and of themselves, as they do not bring substantial rewards from the central government. Meanwhile, the improvement in the looks of buses tends to be done in a quick and short-lived style, paid for by the bus operators and enforced by the city government only during the title application period. Neither the titles nor the beautification projects provide any clear long-term benefits to bus companies themselves. They primarily provide reputational benefits to local leaders.

The second visibility project is bus upgrade projects – purchases of extra-contract, new-model buses. Bus firms usually replace their buses every eight to ten years as part of a standard contract. However, since the average tenure for party secretaries and mayors in Chinese cities is less than three and a half years,Footnote ⁹ they sometimes impose extra-contract demands on bus firms to upgrade the entire bus fleet before their terms expire.

There are two major kinds of upgrades: new-energy buses and “aesthetically pleasing” buses. New-energy buses run on fuel such as liquefied petroleum gas (LPG), liquefied natural gas, or electricity, instead of diesel or petrol. “Aesthetically pleasing” buses are those that look nice (or even “cute”), such as double-deckers, retro buses, and trolley buses. Purchase of these models is usually not specified in contracts.

Upgrade projects are often associated with other city level projects such as “beautiful Zunyi” or “clean Guangzhou” to showcase local leaders’ vision for tourism development or environmental preservation. The benefits of these upgrades are often unclear to everyone except the local government officials. While clean-energy buses could in theory lead to improvements in air quality, the reality is often less clear-cut. For example, LPG fueled buses became popular in the late 1990s for reducing black smoke, but by the mid 2000s, cities began to discover that LPG was producing more pollutants than diesel because of the prevalent use of cheap and ill-suited bus engines incompatible with LPG. Chinese cities that adopted LPG buses have subsequently abandoned them and moved on to other types of fuel.Footnote ¹⁰ Another example is the more recent shift to electric buses without the city investing in charging stations, so bus firms have to revert to traditional fuels and let their new electric buses sit idle.Footnote ¹¹

Meanwhile, these upgraded buses are typically at least twice as expensive as regular buses. Without financial assistance from local governments, firms find these extra-contract purchases very difficult to afford. Even state-owned enterprises (SOEs) with soft budget constraints struggle with ad-hoc upgrades. In one conversation I observed between a bus manager at an SOE and a senior official from the local Fiscal Department regarding a required purchase of a batch of “cute” double-deckers, the SOE manager failed to persuade the official to subsidize the purchase, and the manager said to the official that “the city needs an image, fine, we support that, no problem, it’s our duty. But leaders must understand it costs money, please understand our difficulties!”Footnote ¹²

The third type of visibility project is the construction of BRT, a system of enclosed, dedicated bus lanes with priority given to buses at intersections. BRT can “show the vision of leaders in urban planning” and “looks impressive.”Footnote ¹³ Establishing a BRT typically involves a switch from standard 8-meter-long buses to 12- to 18-meter buses, to increase capacity. It also involves the creation of dedicated lanes servicing new bus routes in the most populous area of the city, to relieve congestion. Chinese cities that adopted BRT have advocated it as a “citizen life quality project” (minsheng gongcheng). BRT construction is entirely a city-level decision that does not require upper-level government approval (unlike subway construction), and is solely financed by the city government.

However, BRTs in Chinese cities are often built quickly within a leader’s term without careful planning, with a construction period between eight months and one year, resulting in incomplete BRT lanes, poorly designed intersections between BRT lanes and regular lanes, and BRT lanes in less populated areas. In many cities, BRTs caused more congestion and traffic accidents than before. Besides this, several cities that built BRTs already had subway systems in place, making the need for BRTs unclear. By June 2017, of the twenty-three cities that had built BRTs, fourteen cities experienced worse congestion, and five cities have subsequently demolished their BRT systems after a few years.Footnote ¹⁴

From the bus firms’ perspective, BRTs are very expensive undertakings. They overlap with many existing bus routes, sometimes the most profitable ones, yet the firms are required to reroute away from the BRT lines. Bus firms are not involved in the decision-making process to build BRT, so they rarely have a chance to negotiate on routing arrangements. Therefore, out of all the visibility projects in the bus sector, firms resist BRTs the hardest. BRTs not only require firms to invest in new and expensive buses, but also cause them to lose many of their most profitable routes.

All three visibility projects – bus beautification, bus upgrades, and BRT – impose significant costs on bus firms but do not bring them obvious benefits. Particularly because local governments strictly maintain inexpensive bus fares as “people are used to low bus fares, otherwise they will be unhappy,”Footnote ¹⁵ bus firms cannot recoup their investments through higher fares. These costs imposed on bus firms are, in effect, corporate contributions to visibility projects and selectoral campaigns of local officials, and are often coerced as firms usually strongly resist them. As a result, state–business relations deteriorate in the sector quickly, and in many cities, local governments eventually chose to force out private firms and replace them with local SOEs to more conveniently launch visibility projects. Before I present qualitative evidence of this process, I use a new dataset to examine the frequency of these visibility projects in Chinese cities and their association with firms’ outcomes.

Effect of Visibility Projects on Private Firms’ Survival

To examine the frequency of visibility projects and its relationship to firm outcomes in China’s urban bus sector, I construct an original dataset covering 233 Chinese cities that privatized urban buses at some point before 2016.Footnote ¹⁶ The dataset contains information about the timing of privatization in any city where it occurred between 1996 to 2016, as well as the timing of the three types of visibility projects. In addition, the dataset records all instances where city governments deprivatized local bus services, the ultimate negative outcome for private firms. Finally, the dataset includes several variables on the characteristics of cities and their public transit systems, as well as leadership tenure. All data are compiled from Chinese newspapers, city government documents, and government yearbooks.

Visibility projects were common. A quarter of cities launched bus beautification projects at least once during the study period, 58 percent launched bus upgrade projects, and 9 percent built BRT. I use a Cox proportional-hazards model to test whether these projects are related to eventual deprivatization. The proportional-hazards model takes the sequencing of annual observations into account and estimates the cumulative probability of deprivatization.

I follow the data setup procedure in Hosmer et al. (Reference Hosmer, Lemeshow and May2008) to account for left truncation, delayed entry, and right censoring of the outcome variable. Cities enter the analysis when they privatize their bus sector, at which point they become at risk of deprivatization. Cities exit the analysis when they deprivatize. These data are left truncated because forty-one cities became at risk – privatized – before 1996. There is also delayed entry for 169 cities in which the bus sector was privatized after 1996. No city experienced deprivatization prior to 1996, therefore data are not left censored. The data are right censored because forty-four cities had not yet deprivatized their buses at the end of the observation period. The process in Hosmer et al. (Reference Hosmer, Lemeshow and May2008) ensures that the proportional-hazards model produces unbiased estimates under these conditions.

I model the cumulative probability of deprivatization – the state-initiated procurement of all private bus firms in a city – as a function of the city’s history of visibility projects. The dependent variable, “Duration of Private Ownership,” is a count of years elapsed between the foundation of the first private bus firm in a city and the start of deprivatization of all private bus firms in a city. Only ownership change that targets all the private bus firms in a city, rather than selecting a number of them, is coded as deprivatization. If individual private firms chose to leave the sector without government pressure, then it is not coded as deprivatization, which makes this measure conservative. Deprivatization is coded as 0 if it does not happen in a year, and 1 for the year it starts.

I hypothesize that contributions to visibility projects in the urban bus sector will increase the risk of deprivatization of the sector, since they are not in the interests of private firms. To avoid accidentally registering projects that are not visibility projects, I only include projects that are obviously initiated by the city government and involve all bus firms in a city.

• H1: Bus beautification projects will raise the hazard rate of deprivatization of the urban bus sector.

Beautification project is a binary variable. A value of 1 indicates an announcement to enter one of the following national city title competitions: National Civilized City, held every three years; and National Public Transportation City, held every year. A value of 0 indicates the city makes no such announcement.

• H2: Bus upgrade projects will raise the hazard rate of deprivatization of the urban bus sector.

Upgrade project is a binary variable. A value of 1 indicates a news report or government announcement of a government-initiated extra-contract purchase of aesthetically pleasing buses or new-energy buses. Absent such an announcement, a value of 0 is recorded.

Regular replacement of aging buses is not included in this measure. For example, if a news story mentions a bus firm is replacing buses at the end of their life, this is coded as 0. If a news story mentions nothing about bus replacement, but simply introduction of new-model buses into the city, it is coded as 1. Only purchases that are obviously state-initiated are included.

• H3: Major infrastructure projects that will impose extra costs on the bus firms will raise the hazard rate of deprivatization of the urban bus sector.

BRT is also a binary variable, with a value of 1 for city/years where BRT construction was announced, and 0 when no such announcement was made. Only announcements of actual BRT construction are coded as 1, while “plans” or “prospects” for BRT is coded 0, since the threat to the bus firms is not yet present or certain.

Any sectoral features affecting the risk of deprivatization (e.g. asset mobility and incomplete contract problems) are automatically controlled for by the single-sector design. However, I do have to condition on factors that potentially affect the government’s decision to deprivatize.

Fiscal Health is the ratio of government expenditures to revenues, which provides a measure of the state’s fiscal capacity to deprivatize the bus firms. After all, fiscal constraints are what motivated the original push to privatize. A lower ratio signals greater fiscal health and that the city has more resources to carry out deprivatization.

Buses per 1,000 people and paved roads per capita capture the fact that larger, denser bus networks present a greater regulatory challenge. Regulatory success is an important factor to the success of privatization (Hart, Shleifer, and Vishny Reference Hart, Shleifer and Vishny1997; Gerard Reference Gerard2008), and Chinese local officials in interviews constantly mention how large systems of urban buses are “hard to manage.” It is possible that deprivatization was a state attempt to compensate for the difficulty in regulating bus services, and if this were true, we should expect deprivatization to be more likely in cities with larger and denser transit systems. I also measure the Routine bus replacement by the annual replacement in number of buses as a percentage of total buses, to take into consideration the regular activity of replacing buses. I include GDP per capita to control for the general wealth level of a city.

Political leaders’ tenure – the term year of a city’s party secretary and mayor – are also included to capture the potential influence of political cycles on local budgeting decisions (Guo Reference Guo2009) that could affect the decision of deprivatization.

I use the Cox proportional-hazards model to test the hypotheses. All models use robust variance estimators to adjust for potential correlation among outcomes from the same city. Model specification 1 presents analysis with the key independent variables. Models 2 and 3 add different control variables to the model.Footnote ¹⁷

As predicted, all visibility projects increase the hazard rate of deprivatization of a city’s urban bus sector over time. Compared with cities that never launched these projects, launching BRT increases the probability of deprivatization by 6.7 times, bus upgrade projects by 4.2 times, and bus beautification projects by 3.8 times (Figure 4.2, model 3).

Figure 4.2 Coefficient plots, Cox proportional-hazards model

As Figure 4.3 shows, the predicted probability of private firm survival declines over time after each wave of visibility projects. Holding all other variables at their mean value, the probability of the bus sector remaining private declined much more quickly in cities that launched visibility projects. For example, only 78 percent of cities will still have a private bus sector within five years after a bus beautification project, compared with 94 percent of cities that did not launch the same project. After ten years, the difference increases to 38 percent versus 76 percent, and after fifteen years it increases to 12 percent versus 52 percent. This pattern of increased deprivatization following visibility projects is consistent across the different project types. This association is even stronger for BRT and upgrade projects, where launching BRT will certainly lead to deprivatization within five years. This is most likely because of the higher costs associated with these projects. These findings are also robust to the use of the nonparametric Kaplan–Meier estimator.Footnote ¹⁸

Figure 4.3 Effects of visibility projects on deprivatization, Cox predicted versus Kaplan–Meier observed survival probability

We would expect, moreover, that cities that launched several waves of visibility projects within a short time frame deprivatized sooner than cities that launched the same number of visibility projects over a longer period. Frequent visibility projects mean rapidly accumulating expenditures and debt for private firms, which in turn increases the risk of deprivatization. Here, I remove BRT projects from the analysis because launching BRT will certainly lead to deprivatization, so I only need to test whether increasing frequency in the other two types of visibility projects will increase the risk of deprivatization. Indeed, Figure 4.4 shows evidence for this “dosage frequency” effect.Footnote ¹⁹ The predicted probability of firm survival declines more rapidly with each additional visibility project launched in the preceding three-year period.Footnote ²⁰ The chance that the bus sector remains private after five years is 98 percent for cities that launched no projects in the past three years, 92 percent for cities that launched one, 79 percent for cities with two, 61 percent for cities with three, 39 percent for cities with four, and 18 percent for cities with five.

Figure 4.4 Predicted survival probability of private firms by visibility project frequency

Clearly, frequent visibility projects lead to faster deprivatization of the sector. This could be because visibility projects increase costs for private firms, which could result in stronger firm resistance to future visibility projects, and eventually to politically expedient deprivatization.

In this chapter, I show the correlation between visibility projects and the survival of private firms across Chinese cities. Once visibility projects are launched, private firms exit the urban bus sector. The more visibility projects are launched and the more frequently they are launched, the quicker private firms exit the sector. But how exactly do visibility projects lead to private firms’ exit, and how do these exits happen? In Chapter 5, I use a comparative case study with process tracing and in-depth interviews to show the mechanism of how visibility projects lead to deprivatization of the urban bus sector.