![Sample our Law journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5942/TOC-Subject-Sample-2021-Law.jpg"})
Abstract
Recent discussions on the natural resource curse theory have expanded from looking at economic and sociopolitical developments to focusing on the impact of natural resources on the spread of, and deaths from, infectious diseases. However, consensus on a link between natural resources and infectious diseases rarely exists, and empirical results are mixed at best. This paper attempts to re-explore such a link by focusing on malaria, a major infectious disease. We argue that in resource-rich countries the reluctance of governments to invest in human capital, rampant corruption and weakened state capacity, and inferior hygiene conditions in mining and drilling areas lead to higher numbers of cases of malaria. To provide empirical support, we apply different natural resource measures, and examine their impacts on the number of cases of infection and death from malaria for the period 2000–2014. Statistical results largely confirm our observations that natural resource abundance is positively associated with a higher number of incidences of and deaths from malaria. These results hold with alternative malaria and resource indicators, and model specifications. The results also have policy implications for malaria control, global public health, and natural resource management.
Notes
Earlier version of this paper was presented at the 2017 Annual Meeting of the Taiwanese Political Science Association, Taipei, Taiwan. The authors would like to thank Jaw-ling Joanne Chang, Eric Chiou, Chia-yi Lee, Howard Liu, Szu-Ning Ping, Tsung-han Tsai, Yi-ting Wang, conference attendees and three anonymous reviewers for helpful comments. This project has been supported by the Ministry of Science and Technology (MoST) of Taiwan [Grant No. MOST 106-2410-H-004-103]. We take the sole responsibility of all errors in this paper.
1 Of six WHO regions (Africa, Americas, South-East Asia, Europe, Eastern Mediterranean, and Western Pacific), the average value of malaria infected case is 35,675,000 with a standard deviation of 75,000,779. TB’s two respective values are 1,731,667 and 1,740,071, and HIV/AIDS’ are 6,110,000 and 9,619,220, respectively (calculated based on CitationWHO, 2017).
2 In this paper, the terms “resource-rich” and “resource-poor” mean more than just the degree of natural resource abundance. More importantly, they denote the degree to which a country relies on natural resources for economic production. While “resource-rich” means heavier dependence in terms of degree, “resource-poor” means less or even no dependence.
3 A further description of total natural resources rents is introduced in Section 3’s Independent variable subsection.
4 There is no denying that in some resource-rich countries, political leaders spend their resource dollars smartly, like in Saudi Arabia and Venezuela, to sustain political life by satisfying people’s needs, such as social welfare, free education, and free healthcare. Yet, overall, more resource-poor countries are erased from the malaria list than resource-rich countries, as we have already seen in .
5 This is also what CitationRoss (2001 called the “taxation effect,” that, by avoiding taxing people or reducing their tax burden intentionally or unintentionally because of oil revenues, the general public will be less likely to request accountability from their government. CitationLevi (1989); CitationMcDonald (2007) reach similar observations on taxation and government accountability.
6 We use three governance indicators: WorldGovernance Indicators(WGI, CitationKraay et al., 2010), Corruption Perceptions Index (CPI, CitationTransparency International, 2017), and State Fragility Index (SFI, CitationMarshall & Elzinga-Marshall 2017). For WGI, this is a score between ±2:5, and higher values denote more political capability. We take the average value of six subcategories of WGI: political accountability, political stability, regulatory quality, rule of law, control of corruption, and government effectiveness to generate the overall WGI score. SFI uses a 0–25 ordinal scale, and higher values denote more fragility. Finally, CPI uses a scale of 0–100, and higher values show less corruption. Because higher CPI and WGI values denote more cleanliness, and more political capability, respectively, to make interpretations consistent, we converted the SFI score to make higher SFI values denote less fragile.
7 The rationale behind widespread embezzlement and fraud with foreign aid happening in Africa is that most global funding goes to Africa. In other regions, like Asia and South and Central America, central governments take the main responsibility for funding malaria elimination programs (CitationPigott, Atun, Moyes, Hay, & Gething, 2012).
8 CitationWorld Bank (2018c) also reports malaria data, namely incidence of malaria per 1,000 people at risk. Yet this dataset only has 258 data points, which is roughly one-fifth of the datasets we used in this paper. We decided not to use this dataset.
9 As renowned resource expert CitationRoss (2015) mentioned, there is no single best indicator that can measure or approximate natural resources directly. Therefore, we have natural resource depletion as an alternative indicator of natural resources to examine the relationship between malaria cases and natural resources.
10 Conventional practice suggests the application of the Hausman test to decide the choice between the RE and the fixed effects (FE) model. Yet scholars do not suggest this test as they simply view it as the standard selection of models, or as the endorsement of the FE model if failing to reject the null hypothesis (CitationBaltagi, 2008; CitationClark & Linzer, 2014).
11 We also took the issue of serial correlation into consideration. Yet, as CitationBaltagi (2013) argues, this is only the case when the time period is sufficiently long, around 20 years or more. Since our investigation period is 15 years, the influence of autocorrelation should not be a serious issue.
12 We conducted the Breusch-Pagan Lagrange Multiplier (LM) test to decide whether the RE or OLS model should be selected as the main one. The null hypothesis of the LM test is that no variances exist across observations. The test result gave us the probability > chibar(01) =0.0000. Therefore, we can reject the null hypothesis that there is no variance across units, and therefore the selection of the RE model is preferred.