Clientelism and Forest loss in a Macro-Comparative perspective

Abstract

For several decades, cross-national scholars have aimed to understand why democracy tends to be related to increased forest loss, despite theory suggesting the exact opposite directional relationship. Recently, Sanford (2021) finds that closer elections in democratic nations tend to increase forest loss. The author attributes this finding to be the result of clientelism or the targeted distribution of goods, services, jobs, and money in exchange for the political support of a candidate. However, we are not aware of any cross-national research that considers if higher levels of clientelism are related to increased forest loss in low- and middle-income nations. To fill this gap, we analyze data for 80 low- and middle-income nations using a two-stage instrumental variable regression model. We find that higher levels of clientelism correspond with increased forest loss after controlling for various economic, political, and demographic factors hypothesized to explain it.

Keywords:

• Forest loss
• cross-national
• clientelism
• democracy

Notes

1 Our sample includes the following 81 low- and middle-income nations from the World Bank’s (2016) classification of nations by income level and listwise deletion of missing data. They are Albania, Argentina, Azerbaijan, Bangladesh, Belarus, Benin, Bolivia, Botswana, Brazil, Bulgaria, Burundi, Cabo Verde, Cambodia, Cameroon, Central African, Republic, China, Colombia, Comoros, Costa Rica, Cote d'Ivoire, Cuba, Ecuador, El Salvador, Eritrea, Ethiopia, Fiji, Gabon, Georgia, Ghana, Guatemala, Guinea, Guyana, Honduras, Hungary, India, Indonesia, Jamaica, Kazakhstan, Kenya, Kyrgyz Republic, Lesotho, Madagascar, Malawi, Malaysia, Maldives, Mali, Mauritius, Mexico, Moldova, Mongolia, Mozambique, Nepal, Nicaragua, Nigeria, Pakistan, Panama, Papua New Guinea, Paraguay, Peru, Philippines, Romania, Russian, Federation, Sao Tome and Principe, Seychelles, Sierra Leone, South Africa, Sri Lanka, Sudan, Suriname, Swaziland, Tajikistan, Tanzania, Thailand, Togo, Turkmenistan, Uganda, Ukraine, Vanuatu, Venezuela, Vietnam, and Zimbabwe.

2 In the first stage of our models, we find that the two instrumental variables predict significant variation in forest loss in the expected directions. This is not surprising given the results of the diagnostics discussed. None of the independent variables are related to clientelism at the p < .05 level for a one-tailed test.

3 There is renewed interest in how foreign investment impacts the natural environment using cross-national data. For example, Mejia (2022a) and Mejia (2022b) find they are related to increased carbon dioxide emissions. We include total inward stocks and flows of foreign direct investment as a percentage of gross domestic product. The data may be obtained from the United Nations (2021). The coefficients for these variables fail to reach a level of statistical significance. This finding may be the result of not considering foreign investment in the primary sector—see Jorgenson (2008). We do not do so because there are limited data available for this measure.

4 We include other measures in our model to demonstrate the reliability of the findings. We examine the impact of trade as a percentage of gross domestic product and debt service as a percentage of total exports of goods and services. The coefficients for these variables do not reach a level of statistical significance.

5 We include Freedom House’s (2005) political rights and civil liberties measures as alternative independent variables to assess the impact of democracy on forests. The political rights measure reflects whether a nation is governed by democratically elected representatives and has fair, open, and inclusive elections (Freedom House 2005). The civil liberties variable reflects the extent to which a nation has freedom of the press, freedom of assembly, general personal freedom, freedom of private organizations, and freedom of private property (Freedom House 2005). The variables are measured on the following seven-point scale: free (1–2), partially free (3–5), and not free (6–7). The coefficients for political rights and civil liberties do not reach a level of statistical significance when included into the models separately or together.

6 We do not find that international non-governmental organizations are related to less forest loss. This is not surprising as cross-national research recently finds various interaction effects involving this measure. For example, Henderson and Shorette (2017) find that international non-governmental organizations decrease the effect of palm oil production on forests in peripheral nations. Sommer (2021) finds that high levels of interference from other nations mitigate the ability of international non-governmental organizations to protect forests. Tasmim et al. (2020) find that international non-governmental organizations enhance the ability of domestic non-governmental organizations to reduce forest loss.

7 We include the percentage of a country’s population in urban areas for 2000 to consider how another demographic factor may impact forests. The data come from the World Bank (2016). Rudel (2017) finds that urbanization is related to increased forest loss as cities are expanding into previously remote areas. We do not find support for this line of reasoning.