16
J. Swart and L. Brinkmann
the square term in the subsequent specifications to allow for a linear relationship
between the environmental degradation indicator and complexity.
Column (2) reports the estimation results of Eq. (2), which includes GDP per
capita and its square term as additional explanatory variables. The ECI estimate is now
statistically significant and negative, indicating that a rise in economic complexity
leads to a reduction of the amount of solid waste generated daily per person. For
concreteness, individuals generate an average of 3.63 g less waste per day if the
complexity index increases by one unit. This finding shows that waste generation
has a negative relation with economic complexity, which suggests that individuals
generate less waste in more complex municipalities. An intuitive reasoning is that
in more economic complex municipalities, there is a higher percentage of tertiary
sectors which generate less waste than manufacturing sectors. Another reason could
be that more complex economies have more knowledge about how to make more
efficient use of resources and materials, leading to less waste.
Both GDP per capita and its square term are significantly different from zero.
With the square term bearing a negative sign, the relationship between solid waste
generation and income per capita is inverted U-shaped. This finding supports the
existence of a waste generation EKC for the state of São Paulo.
The two significant relationships found in specification (2) are robust to the inclusion of individual controls and the complete set of variables. Population density and
the time trend are unrelated to solid waste generation. In contrast, trade openness
has a significant and negative impact on the dependent variable, suggesting that
individuals generate less solid waste with rising trade.
Overall, our estimation results provide evidence that rising economic complexity
and trade openness have a significant negative impact on the amount of solid waste
generated per person per day in the municipalities of the state of São Paulo. In line
with the EKC concept, our estimates further confirm the existence of an inverted
U-shaped relationship between solid waste generation and economic development.
The turning point, i.e. the level of per capita income at which the detrimental effect of
economic development on the environment reverses, lies between 59,120 and 61,499
constant 2000 Brazilian Reais. These values lie within the income per capita range
in our sample (see Table 9 in the Appendix).
4.2 Deforestation
Table 2 reports the first differences estimates for our analysis on the relationship
between deforestation and economic complexity in municipalities in the Amazon
basin region between 2002 and 2014. We find no evidence of a relationship between
economic complexity and deforestation. Our estimates provide evidence, instead, for
an inverted U-shaped relationship between deforestation rates and economic development. GDP per capita and its square term yield statistically significant estimates
in all specifications. Deforestation rates soar with rising income per capita, but this
J. Swart and L. Brinkmann
the square term in the subsequent specifications to allow for a linear relationship
between the environmental degradation indicator and complexity.
Column (2) reports the estimation results of Eq. (2), which includes GDP per
capita and its square term as additional explanatory variables. The ECI estimate is now
statistically significant and negative, indicating that a rise in economic complexity
leads to a reduction of the amount of solid waste generated daily per person. For
concreteness, individuals generate an average of 3.63 g less waste per day if the
complexity index increases by one unit. This finding shows that waste generation
has a negative relation with economic complexity, which suggests that individuals
generate less waste in more complex municipalities. An intuitive reasoning is that
in more economic complex municipalities, there is a higher percentage of tertiary
sectors which generate less waste than manufacturing sectors. Another reason could
be that more complex economies have more knowledge about how to make more
efficient use of resources and materials, leading to less waste.
Both GDP per capita and its square term are significantly different from zero.
With the square term bearing a negative sign, the relationship between solid waste
generation and income per capita is inverted U-shaped. This finding supports the
existence of a waste generation EKC for the state of São Paulo.
The two significant relationships found in specification (2) are robust to the inclusion of individual controls and the complete set of variables. Population density and
the time trend are unrelated to solid waste generation. In contrast, trade openness
has a significant and negative impact on the dependent variable, suggesting that
individuals generate less solid waste with rising trade.
Overall, our estimation results provide evidence that rising economic complexity
and trade openness have a significant negative impact on the amount of solid waste
generated per person per day in the municipalities of the state of São Paulo. In line
with the EKC concept, our estimates further confirm the existence of an inverted
U-shaped relationship between solid waste generation and economic development.
The turning point, i.e. the level of per capita income at which the detrimental effect of
economic development on the environment reverses, lies between 59,120 and 61,499
constant 2000 Brazilian Reais. These values lie within the income per capita range
in our sample (see Table 9 in the Appendix).
4.2 Deforestation
Table 2 reports the first differences estimates for our analysis on the relationship
between deforestation and economic complexity in municipalities in the Amazon
basin region between 2002 and 2014. We find no evidence of a relationship between
economic complexity and deforestation. Our estimates provide evidence, instead, for
an inverted U-shaped relationship between deforestation rates and economic development. GDP per capita and its square term yield statistically significant estimates
in all specifications. Deforestation rates soar with rising income per capita, but this