Academic journal article Political Research Quarterly

A Chinese Popularity Function: Sources of Government Support

Academic journal article Political Research Quarterly

A Chinese Popularity Function: Sources of Government Support

Article excerpt

(ProQuest: ... denotes formulae omitted.)

The concept of the "vote-popularity function" (VP-function) was fully developed by Nannestad and Paldam (1994) in a seminal, much-cited paper. (Currently, a special issue of Public Choice featuring this work is underway.) As they argued, "The VP-function explains support for the government as a function of economic and political outcomes" (Nannestad and Paldam 1994, 213). In their review, they counted some two hundred relevant studies, from democracies around the world. Since that time, the number of such studies has grown to at least five hundred, according to a contemporary survey (Lewis- Beck and Stegmaier, forthcoming).

Thus, this field of research has greatly increased. But all these studies, with vote or popularity as the dependent variable, are on democratic systems of some form. It must seem strange, then, to propose a Chinese popularity func- tion, because that system lacks the democratic fundamen- tals. However, the theoretical insight for this idea was planted, in fact, by Nannestad and Paldam themselves, in their "note on dictators and regime shifts-the concept of an indirect VP-function" (1994, 236). They observe that such countries seldom publish government popularity data but "If one can find a systematic government reac- tion . . . one can estimate an indirect VP-function" (Nannestad and Paldam 1994, 236). Fortunately, we do have available such systematic popularity data on China, as elaborated below.

Chinese citizens, as we know, do not vote in national elections. But that does not mean they are without other collective modes of national political expression. A lead- ing manifestation of their voice comes from public opinion, expressed frequently in informal ways. Several studies show how opinions expressed online, in petitions, through village elections, and in protests and demonstra- tions affect government decision making (O'Brien 2008; Shirk 2011; Tang and Iyengar 2012). More importantly for our purposes, the public is also being systematically followed in formal sample surveys. Chinese public opin- ion via sample survey has come under increasing scru- tiny, in a growing number of mass-based studies (see the following outstanding examples: Chen 2004; Dalton and Shin 2007; Jennings and Chen 2008; Shi 1997; Tang 2005). The article at hand continues in this tradition, through exploration of a major and recent national prob- ability survey of the Chinese adult population. However, it departs from these other studies, in that it aims to develop a popularity function. It tries to answer the ques- tion-What drives the Chinese citizenry to increase (or decrease) their support for the national government?

To begin, we consider VP-function theory, and evalu- ate its applicability (or inapplicability) to the Chinese case. Then, we describe the data set, focusing especially on the univariate characteristics of the popularity variable. Next, we offer a model of Chinese government support. The organization of the model, as a VP-function, emphasizes gross political and economic forces. Furthermore, because the data are cross-sectional, socio- logical forces are also included. The estimation strategy, a block-recursive one, takes inspiration from a Michigan- style funnel of causality, building out from more exoge- nous to less exogenous independent variables. As shall be seen, an empirically and theoretically sound model of national government support in China appears within reach.

The VP-Function: A Theory of Government Support

The VP-function, as developed by Nannestad and Paldam (1994, 213), holds government support (as measured by popularity or vote) to be determined by basic political and economic attitudes and performance. Here, in words, it is expressed in the function

... (1)

Because the Chinese observations on popularity are cross-sectional, we also add to the function basic sociode- mographic conditions, which vary considerably in a cross-section, unlike their essentially constant state in the typical aggregate time-series popularity function. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.