Geography of government support in China: a case study of college students

ABSTRACT

Scholars have debated why people support the Chinese government, but few have studied the spatial pattern of such support. In this article, drawing on a nationwide survey of college students in 2017 (N = 21,674), cumulative link mixed models are used to study the factors accounting for government support in China at individual, prefectural and provincial levels. The results show that competing economic–nationalist, institutional and cultural theories of government support in China all contain elements of truth. However, students in this sample support different levels of the government for different reasons, and these factors vary across places and by geographical scale. In general, economic performance mainly explains support for local government and political and ideological considerations mainly explain support for central government.

摘要

中国政府支持度的地理特征: 以大学生为例。 Area Development and Policy. 一直以来学者们对人们为何支持中国政府争论不休，但鲜有人对于这种支持的空间格局进行研究。本文借鉴2017年全国大学生调查（N = 21,674），使用累积链接混合模型，从个人、地市级、省级三个层面来研究对中国政府支持度造成影响的因素。研究结果表明，在中国，经济民族主义理论、制度理论、文化理论都对政府支持度具有一定解释力。并且，案例研究显示学生们出于不同原因而支持不同级别的政府，这些因素因地点和层级变化而有所不同。总体而言，经济表现主要诠释了大学生对地方政府的支持，而政治和意识形态上的考虑主要诠释了他们对中央政府的支持。

Geografía del apoyo al Gobierno en China: el ejemplo de estudiantes universitarios. Area Development and Policy. Los académicos han debatido cuáles son los motivos del apoyo al Gobierno por parte del pueblo chino, pero pocos han estudiado el patrón espacial de ese apoyo. En este artículo, a partir de una encuesta nacional a estudiantes universitarios en 2017 (N = 21.674), se utilizan modelos mixtos de enlaces acumulados para estudiar los factores que explican el apoyo al Gobierno de China a nivel individual, provincial y territorial. Los resultados indican que todas las teorías contrapuestas de ámbito económico-nacionalista, institucional y cultural del apoyo al Gobierno de China tienen algo de verdad. Sin embargo, los estudiantes en esta muestra apoyan diferentes niveles del Gobierno por diferentes motivos, y estos factores varían según el lugar y la escala geográfica. En general, el rendimiento económico explica sobre todo el apoyo a administraciones locales, y las consideraciones políticas e ideológicas explican principalmente el apoyo al Gobierno central.

География поддержки государства в Китае: обследование студентов вузов. Area Development and Policy. Ученые обсуждают, почему люди поддерживают китайское правительство, но мало кто изучал пространственное распределение такой поддержки. В этой статье, опираясь на общенациональный опрос студентов вузов в 2017 году (N = 21 674), смешанные модели кумулятивных связей используются для изучения факторов, связанных с поддержкой государства в Китае на индивидуальном, окружном и провинциальном уровнях. Результаты показывают, что конкурирующие экономико-националистические, институциональные и культурные теории государственной поддержки в Китае содержат элементы истины. Однако студенты в этой выборке поддерживают разные уровни власти по разным причинам, и эти факторы варьируются в зависимости от места и географического масштаба. В целом экономические показатели в основном объясняют поддержку местных властей, а политические и идеологические соображения в основном объясняют поддержку центрального правительства.

KEYWORDS:

• government support
• college students
• public opinion
• cumulative link mixed model
• China

JEL:

• F52
• O17
• P25
• P26

关键词:

• 政府支持度
• 大学生
• 公众舆论
• 累积链接混合模型
• 中国

PALABRAS CLAVE:

• apoyo al Gobierno
• estudiantes universitarios
• opinión pública
• modelo mixto de enlaces acumulados
• China

КЛЮЧЕВЫЕ СЛОВА:

• государственная поддержка
• студенты вузов
• общественное мнение
• смешанная модель кумулятивной связи
• Китай

ACKNOWLEDGEMENT

The authors are grateful to Ernst Linder, Linyuan Li and Lawrence C. Reardon at the University of New Hampshire for their kind advice on this article. The authors also thank the three referees and editor for their invaluable comments. Any errors remain the authors’ own. There was no funding for this research.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. In China, prefecture is the level of administration between province (sheng) and county (xian). In Chinese, a prefecture is also called ‘city’ (shi).

2. We are fully aware of the nuanced relationship between political trust and political support. Following prior researchers (Shi, 2015; Zhong, 2014), we treat them interchangeably here to simplify our analysis. For a thorough investigation of the relationship between political trust and political support, see Hetherington (1998).

3. By September 2017, there were 334 prefecture-level units in mainland China. We combine some of the prefectural-level units directly administered by the provincial government in four provinces. They are 15 units in Hainan (Baoting Li and Miao Autonomous County, Dingan County, Ledong Li Autonomous County, Changjiang Li Autonomous County, Lingao County, Lingshui Li Autonomous County, Qiongzhong Li and Miao Autonomous County, Danzhou City, Baisha Li Autonomous County, Wanning City, Tunchang County, Wenchang County, Dongfang City, Qionghai City and Chengmai County), one in Henan (Jiyuan City), three in Hubei (Xiantao City, Qianjiang City and Tianmen City) and six in Xinjiang (Alar City, Tumushuk City, Shihezi City, Beitun City, Tiemenguan City and Wujiaqu City). These counties are thus grouped into four prefecture-level units. Due to data availability, we do not include Ali in Tibet, Sansha, and Danzhou in Hainan. For convenience, we also code the four municipalities (Beijing, Tianjin, Shanghai and Chongqing) and two special administrative regions (Hong Kong and Macau) as prefecture-level cities. Therefore, we end up having 334 + 4 + 2 + 4–3 = 341 prefecture-level units.

4. We use the terms ‘effects’ and ‘random effects’ purely in the statistical sense. Since our research intends to be exploratory, we do not imply that we have proven causation by using these words.

5. We do not include regional variables directly in our multilevel models for two reasons. First, we have five provincial-level variables but only three prefectural-level variables available. In the multilevel models, we want to keep the regional variables consistent across scales. Second, since the spatial complexity in our analysis is 3 (individual, prefecture and province) by 2 (home and university), including group variables directly in the model would make the results difficult to explain. Using the coefficients of a first model as regressors in a second model has become common in the social sciences, but these second regressions often have additional requirements because the assumptions of the linear regression are not met. Therefore, we use the method first developed by Hanushek (1974) to run weighted Pearson’s correlations between the REs and group variables.