The effects of mining presence on inequality, labor income, and poverty: evidence from Peru

Abstract

This paper studies the effects of mining intensity and presence on Peru’s mining districts’ welfare from 2004 to 2019. A pooled cross-section regression is used which is constructed from different sources and two sets of comparisons are made: the first compare districts with and without mining presence within mining provinces, and the second compares districts with and without mining presence without the constraint of being within mining provinces. The primary dependent variables included in the model are income inequality, labor income, and poverty rate. In mining districts, inequality has increased, but labor income has increased, and poverty has decreased compared to non-mining districts. However, once control for province-fixed effects and clustered by standard errors at the district level, the significance of inequality is lost, while the impacts on labor income and poverty remain. The transmission mechanisms are human capital, employment, and redistributive policies. Also the mining presence has had positive effects on labor income in other sectors such as construction and commerce; Finally, the labor incomes of unskilled workers increases but not the labor incomes of skilled workers, and it has negatively impacted informal employment.

Appendices

Appendix 1

Tables 13, 14, 15, 16, 17, 18, 19 and 20

Table 13 Descriptive statistics: first comparison group

National Institute of Statistics, Thomson Reuters and Ministry of Mining and Energy, and author’s calculations. Table 13 shows the descriptive statistics of the dependent and independent variables by first comparison group.

Table 14 Descriptive statistics: second comparison group

Appendix 2

In this appendix, I show the results of the effects of the presence of mining companies on the level of education:

$${EDUC}_{i,j,t}=\alpha +{\beta }_{1}{Mining}_{i,j,t}+B{X}_{i,j,t,h}+\gamma {Price}_{t}+\omega {GDP}_{i,j,t}+\theta {Canon}_{i,j,t}+\Gamma +{e}_{j,t}$$

(5)

where ${EDUC}_{i,j,t}$ is the labor income vector that considers people who have attained secondary education or less, and the second one considers people who have attained technical higher education or more. ${Mining}_{j,t}$ is the vector containing the measures of presence and intensity of mines, ${X}_{i,j,t,h}$ is the vector of socioeconomic characteristics of person h in district i within province j in period t. ${Price}_{t}$ is the vector of mineral prices over time t, ${GDP}_{i,j,t}$ is the GDP of district i within province j for period t. ${Canon}_{j,t}$ is the canon received by district i in province j in period t. $\Gamma$ are the fixed effects of the provinces and ${e}_{j,t}$ are the clustered standard errors at the district level. Table 15 presents the results considering the first comparison group.

Table 15 First comparison group

The results in Table 16 show that the mining presence has increased the demand for high school education or less, because the earnings of unskilled workers have increased while the earnings of skilled workers have not been impacted. Table 16 presents the results for the second comparison group.

Table 16 Second comparison group

The results in Table 16 show that the mining presence has increased the demand for secondary education or less, because the labor income of unskilled workers has increased while the labor income of skilled workers has not been impacted. These results are in line with the literature that points out and finds that the presence of resource abundance decelerates the accumulation of education, because local communities prefer to specialize in activities related to the natural resource being extracted, these results are in line with those found by Michaels (2010) for the case of coal in the United States. Aragón and Rud (2013) for the specific case of Yanacocha in Cajamarca, unskilled workers saw their real labor income increase while there was no significant change in the real labor income of skilled workers.

Appendix 3

In this appendix, I show the results of the effects of the presence of mining companies on labor income across sectors, using econometrics as follows:

$${income}_{a,i,j,t,h}=\alpha +{\beta }_{1}{Mining}_{i,j,t}+B{X}_{i,j,t,h}+\gamma {Price}_{t}+\omega {GDP}_{i,j,t}+\theta {Canon}_{i,j,t}+\Gamma +{e}_{j,t}$$

(6)

where ${income}_{a,i,j,t,h}$ is the labor income of person h working in sector a of district i within province j in period t. ${Mining}_{j,t}$ is the vector containing the measures of presence and intensity of mines. ${X}_{i,j,t,h}$ is the vector of socioeconomic characteristics of person h in district i within province j in period t. ${Price}_{t}$ is the vector of mineral prices over time t. ${GDP}_{i,j,t}$ is the GDP of district i within province j in period t. ${Canon}_{j,t}$ is the canon received by district i in province j in period t. $\Gamma$ are the province fixed effects and ${e}_{j,t}$ are the clustered standard errors at the district level. Table 17 presents the results for the first comparison group.

Table 17 First comparison group

The results in Table 17 show that neither agriculture nor manufacturing revenues have been affected by the mining presence. On the other hand, other sectors such as Construction and Commerce have seen their revenues increase due to the mining presence. Table 18 presents the results considering the second comparison Group.

Table 18 Second comparison group

The results in Table 18 confirm those found in Table 18. The mining presence has positively impacted two sectors, which are Construction and Commerce. These results would support the hypothesis of Loayza and Rigolini(2016) and the findings found by Aragón and Rud (2013), the first paper points out that the mining presence could positively impact the local economy through investments made in that locality, these investments are mainly in infrastructure, on the other hand the second paper finds that the presence of Yanacocha boosted local trade through the purchase of locally produced inputs which boosted the increase of both nominal and real labor income. Like Aragón and Rud (2013) I find no evidence that the mining presence has increased labor income in the agricultural sector.

Appendix 4

This appendix estimates the effects of the presence of mining companies on formal and informal employment labor income, this variable has not been included in the model of the main results because it only has data availability from 2011 onwards, while the main results consider the period 2004– 2019. For this I use the following econometric specification:

$${Informality}_{i,j,t}=\alpha +{\beta }_{1}{Mining}_{i,j,t}+B{X}_{i,j,t,h}+\gamma {Price}_{t}+{GDP}_{i,j,t}+\theta {Canon}_{i,j,t}+\Gamma +{e}_{i,j,t}$$

(7)

where ${Sector}_{a,i,j,t,h}$ is the labor income of person h working in the informal or informal sector in the of district i within province j in period t. ${Mining}_{j,t}$ is the vector containing the measures of presence and intensity of mines, ${X}_{i,j,t,h}$ is the vector of socioeconomic characteristics of person h in district i within province j in period t. ${Price}_{t}$ is the vector of mineral prices over time t, ${GDP}_{i,j,t}$ s the GDP of district i within province j in period t. ${Canon}_{j,t}$ is the canon received by district i in province j in period t. $\Gamma$ are the province fixed effects and ${e}_{j,t}$ are the clustered standard errors at the district level. Table 19 presents the results for the first comparison group.

Table 19 First comparison group

The results in Table 19 show that the mining presence has had no effect on either formal or informal employment labor income. On the other hand, I found that the district GDP had negative effects on formal and informal employment labor income. And the mining canon had positive effects on the labor income of informal employment. However, I found that the mining presence had negative effects on informal employment. Table 20 presents the results considering the second comparison group.

Table 20 Second comparison group

The results in Table 20 validate those found in Table 19. The mining presence had no effect on formal and informal employment labor income, but it did reduce informal employment in the mining districts.