Are Autocracies Bad for the Environment? Global Evidence from Two Centuries of Data.

AuthorSinha, Apra

    Increasing concentrations of carbon dioxide (one of the major anthropogenic greenhouse gases [GHGs]) has been argued to be one of the major causes of rising global temperature. The Fifth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC) suggests that if carbon emissions continue at the present rate, then global mean temperature will increase by 4 C or more above pre-industrial levels by the end of the 21st century (Collins et al., 2013). It is widely agreed that the global mean temperature should not be allowed to increase by more than 2 C above pre-industrial levels, in order to minimize the risk posed by climate change (Edenhofer et al., 2014). (1) Achieving the 2 C target implies serious reductions in carbon dioxide emissions. (2) So far, the focus of the literature on maintaining the 2 C target has been centered around renewable sources of energy and improvements in energy efficiency (Grubb, 2014, p. 14).

    Although there is consensus that reducing emissions is of utmost priority, it is unclear which types of political institutions are better able to achieve this. One can argue that the steps suggested by science for reducing emissions are incompatible with the basic tenets of democracy, such as freedom to choose (Burnell, 2012; Fiorino, 2018). Democracies, with private property rights and individual liberty, are more suitable than autocracies for business operations; therefore, democracies might exert more pressure on the environment owing to the higher level of economic activity. Reducing emissions may have short-run economic costs, and incumbent democratic governments may not be willing to risk the effects of these on their chances of re-election. Not only this, the urgent measures required for climate change mitigation may make democracies vulnerable if it is considered that autocracies are better able to protect the environment. These issues are not faced by autocratic governments, and therefore those governments may be better able to implement tough measures against climate change.

    On the other hand, Payne (1995) argued that, in democracies, people are better informed about environmental problems due to press freedom, and can better express their environmental concerns via freedom of speech. People can organize for a better environment (freedom of association) and put pressure on the political establishment to improve environmental conditions. In non-democratic systems the ruling class is under no such pressure to improve environmental standards. Moreover, democracies place more value on the quality of human life, whereas autocracies are preoccupied with preserving power for the autocrat. In democracies the institutions are responsive and politicians must care about public demands in order to improve their chances of being elected. Also, in democracies governments are held accountable for their actions. Our understanding of and possible solutions to negative effects of climate change are evolving; in democracies mistakes in these areas are more likely to be discussed, and public decision-making is expected to be better informed (Burnell, 2012).

    Dasgupta and Maler (1995) argue that political and civil liberties in democracies are instrumental in protecting the environmental resource-base, at least when compared with the absence of such liberties in countries run by authoritarian regimes. Olson (1993) and Deacon (2003, 2009) also suggest that democracies will provide a better environment. Their argument stems from the politico-economic structure of autocracies, wherein a large chunk of productive resources are owned by a few elites who care less about public goods compared to private benefits. Since the imposition of better environmental standards can lower private consumption but the marginal benefit of a better environment (a public good) is the same for all groups, it is likely that autocratic elites would not be in favor of higher environmental standards. Therefore, one can argue that democracies are better positioned to address climate change issues.

    The relation between economic activity and carbon dioxide emissions is usually formulated in terms of the environmental Kuznets curve (EKC). The EKC is a reduced-form relationship between per-capita emissions and per-capita income; it suggests that per-capita carbon dioxide emissions initially increase with increased per-capita income, and then decline after a threshold level of per-capita income, known as the "turning point." Grossman and Krueger (1995) argue that changes in income affect living standards, environmental regulations, technology, and industrial compositions, and thus environmental pollution. It is well known that as per-capita income increases, initially the share of manufacturing in gross domestic product increases, and then, after a certain level of per-capita income is reached, the share of manufacturing declines. This justifies the nonlinear relationship between emissions and per-capita income used in this paper to explore the effect of institutions on emissions. (3)

    Using data from 150 countries, we show that political institutions are important for determining this turning point. Our primary measure of political institutions is Polity 2. Political institutions are expected to work through marginal emission intensity (i.e., the change in per-capita emissions for a unit change in per-capita income). We anticipate that democratic institutions would have more environment-friendly methods of economic activity as there would be greater demand for better environmental conditions. We provide evidence of significant differences in marginal emission intensity; democracies emit less carbon dioxide per capita for a unit increase in per-capita income. Furthermore, this difference is not driven by outsourcing of polluting activities to autocracies. Our results with consumption-based carbon dioxide emissions are similar. This paper makes the following main contributions.

    First, in the EKC literature the impact of institutions on turning point estimates has been criticized due to the possibility of omitted variable bias (Stern 2004; Kaika and Zervas, 2013a, 2013b). We test for this bias using Oster's (2019) test for omitted variable bias. This test is based on the relative importance of possible excluded and included variables. The test in this paper implies that omitted variables must be five or more times more relevant in comparison to the included variables for the coefficient of interest (i.e., interaction of institutions and per-capita income) to be insignificant (this coefficient represents the difference in marginal emission intensity between autocracies and democracies). As we have included the relevant variables used in the literature, we do not believe that omitted variables would be five or more times more relevant in comparison to the included variables. Thus, the test results show that it is unlikely that the significance and the sign of the coefficient associated with interaction of per-capita income and institutions is biased due to omitted variables.

    Second, one can argue that political institutions are an endogenous outcome and are partly caused by emissions--for example, high-emission countries may choose to remain autocratic. This is especially relevant in cases of high emission levels embedded in global trade. Emission-exporting countries (higher emissions from production than embedded in consumption) may choose to remain non-democratic. This will make political institutions endogenous and may bias our results. Moreover, the institution and per-capita income could be correlated with other omitted variables and thus give us biased estimates of the interaction coefficient (institutions and per-capita income). We use one instrument each for these two variables to solve the endogeneity issue. Recent experiences such as the Arab Spring suggest that democratization movements occur in waves. Although there is no consensus behind the start and reversal, the existing literature suggests that these are not driven by economic conditions (Bonhomme and Manresa, 2015). Based on these factors, we use regional waves of democratization and transition to autocracy as an instrument for democracy, similar to Acemoglu et al. (2019). Further, we use mean per-capita income of other countries in the region as an instrument for per-capita income (see Panizza and Jaimovich, 2007; Alesina et al., 2008; Ilzetzki and Vegh, 2008). Using these two instruments we obtain causal estimates of the coefficient of the interaction of institutions and per-capita income and confirm that democracies emit less carbon dioxide per capita for a unit increase in per-capita income. More specifically, democracies emit 80 kg less carbon dioxide per-capita for a USD1,000 increase in per-capita income, keeping everything else same.

    Third, the concern for environment is a recent phenomenon. In 1965, the US president's science advisory committee put forward concerns about greenhouse effects for the first time. In 1970, the first ever Earth Day was held, and in 1975 the geoscientist Wallace Broecker coined the term global warming. We show that the lower marginal emission intensity for democratic countries shows a pattern that coincides with the surge in people's concerns about climate change and intergovernmental initiatives to reduce emissions. Only in recent decades has the marginal emission intensity in democracies been lower than that in autocracies.

    Fourth, we use Regimes of the World data that categorizes countries into four groups: (1) closed autocracy, (2) electoral autocracy, (3) electoral democracy, and (4) liberal democracy. We find that a unit increase in per-capita income leads to significantly lower additional per-capita emissions in democracies compared to autocracies, with a significantly lower turning point in democracies. The turning point in autocracies is more than twice that in democracies. We also show...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT