February 18, 2022

 
## [1] 1
## [1] 1
## [1] 0

Problem definition

Forestry has accounted for the majority of tree cover loss in the United States. We test the hypothesis that this dominant driver of deforestation is associated with rising incomes and growing population. What do these socioeconomic factors indicate about the short-term future of deforestation?

Data

Table 1: Variable definitions and sources
Variable Definition Source
Industrial Production Index that tracks real output for all facilities located in the United States agriculture, forestry, fishing, and hunting logging. Monthly, seasonally adjusted series, Jan. 1972 to Dec. 2021 Board of Governors of the federal Reserve System (US)
Population Active population aged 15 and over of all person in the United States. Monthly, seasonally adjusted series, Jan. 1960 to Dec 2021. Organization for Economic Co-operation and Development
Income Real personal income in billions of chained 2012 dollars. Monthly, seasonally adjusted series, Jan. 1959 to Dec 2021. U.S. Bureau of Economic Analysis

Forecast variable

Methodology

We adjusted for autocorrelation by putting in the lag of production index to predict the dependent variable. We also used first differencing for the dependent and explanatory variables in order to make them stationary. A step dummy was used to adjust for the effects of COVID and a spike dummy was used to adjust for the 2009 recession.

Regression results

Table 2: Models predicting production index
MLR ARIMA ARIMAX
intercept 4.620***
(0.012)
Population Growth 1.010 0.831**
(1.920) (0.392)
Income Growth 0.411 0.220***
(0.419) (0.078)
trend -0.000
(0.000)
2009 Recession dummy -0.359*** -0.090***
(0.136) (0.029)
COVID period dummy -0.167*** -0.260***
(0.032) (0.035)
ar1 -0.234***
(0.040)
ma1 -0.303***
(0.040)
BIC -648.6 -2204.3 -2258.9
Observations 598 598 598
Notes:
  • p<0.10; ** p<0.05; *** p<0.01

Forecasting results

Table 3: Forecast of initial claims per 10,000 persons
MLR ARIMA ARIMAX
Jan 2022 0.381 1.22 0.209
Feb 2022 -0.430 0.00 -0.253
March 2022 -0.629 0.00 -0.369
April 2022 0.620 0.00 0.372
May 2022 0.226 0.00 0.180
June 20222 -0.244 0.00 -0.179
Six-month change -0.075 1.22 -0.041

Conclusion

  • Our best ARIMA model indicates that deforestation last month decreases deforestation this month.
  • Using our best regression model, ARIMAX, the effects of population growth and income growth can be interpreted.
    • A 1% increase in population growth is associated with a .831% increase in industrial production logging holding all variables constant. This is significant at the 5% level.
    • A 1% increase in income growth is associated with a .220% increase in industrial production logging holding all variables constant. This is significant at the 1% level.
    • There was a a negative change in trend of deforestation due to the 2009 recession and COVID. Both changes in slope are significant at the 1% level.
  • Over the entire 6 month forecast, ARIMAX is predicting a .041% decrease in deforestation.

Implications

  • Overall, these findings reveal that population growth and income growth increase deforestation. Population growth is expected to grow between 400 and 500 million in 2060 and net forest loss will increase to 15 million hectares according to the U.S. Department of Agriculture (2016). Forests provide carbon dioxide, climate regulation, air and water purification, flood protection, and wildlife habitat. Without an increase in government regulation, deforestation in the United States could be detrimental if these benefits become absent due to population growth.

References