Does political democracy enhance human development in developing countries? A cross-national analysis

American Journal of Economics and Sociology, The,  April, 2006  by Ming-Chang Tsai

• E-mail
• Print
• Link

Social Spending of the Government and Democracy

State health and educational spending. Surprisingly, state spending on health and education to date has not been included in causal examinations, even in studies explicitly attempting to test HD hypotheses (Dixon and Moon 1987; London and Williams 1988, 1990; Moon and Dixson 1985; Ranis, Stewart, and Ramirez 2000). Because in most DCs state spending on both education and health aims to provide services that can benefit those in low-income groups more than those in the highest income levels, (4) this study uses both indicators to measure the encompassing policy of the state and expects them to (1) positively correlate with democracy and (2) favorably influence HD. Health spending (computed as a percentage of GDP) consists of government spending plus borrowings and grants for all relevant items. Meanwhile, educational spending is calculated as a percentage of GNP, including government spending on all levels of public schools. (5) Because both expenditures remained relatively stable over the period analyzed, a percentage measure is suitable. Regarding educational spending, a measure of spending for primary and secondary schools only would be a more appropriate indicator of state efforts to target the general population rather than privileged groups that are more likely to benefit from universities; however, scholars have warned that disaggregated educational spending in DCs is unreliable and generates bias in causal assessment (Behrman and Rosenzweig 1994). This study thus decided not to use such measures.

Democracy. This study suggests that two institutional features of democracy (majority rule and political contention) require careful measurement. In measuring majority rule in a polity, the POLITY Project Data (PPD) developed by Ted Gurr (2003) provides for detailed measurement in most countries of (1) citizen participation; (2) institutional constraints on the exercise of power; and (3) guarantee of civil liberties to all citizens. The emphasis is on political participation and the civil rights that the masses enjoy in a political system. Since many countries may simultaneously display some "state-centered" characteristics, the PPD also provides a countermeasure of democracy, termed autocracy. This study thus employs a mixed score (democracy minus autocracy), as the PPD does, to measure the level of encompassing democracy. For each country, a weighted average score (ranging from 10 to -10) was calculated for the period of 1985-1994 to construct a lagged independent variable. This design of lagged factors was also applied to other independents and controls.

The contention approach to the measurement of democracy used in this study is recoded from Przeworski et al. (2000), whose measure focuses specifically on the dimension of elite competition. Based on the notion of democracy as involving the contestation of office among different political parties, a democracy must meet the following conditions: (1) the chief executive must be elected; (2) the legislature must be elected; and (3) there must be more than one party (2000: 19-20). Przeworski et al. (2000) documented contestation events during 1950-1990 for 141 countries. This data set also identifies the bureaucratic-authoritarian (B-A) government, which has some internal roles for operating the government and regulating the competence of the chief executive versus the legislature (Przeworski et al. 2000: 32). The final type of regime is monolithic dictatorship, involving neither legislature nor parties. This study used this rich information for individual countries and created a set of dummy variables, with the dictator being the omitted "comparison group." Conceptually, political contention, as this study operationalizes it, is a specific dimension of liberal democracy. But empirically, its potential effect on HD deserves serious research attention. The main disadvantage of this data set is that its coverage ends in 1990, leaving out five years uncovered by our analysis, and thus the results must be reinterpreted accordingly. (6) Przeworski and his co-authors found that their democracy measure is a strong predictor of economic growth, as well as of certain physical well-being indicators. This study expects to find that the set of democracy variables show evidence of democratic regimes enhancing HD.