← PLSC 441 · The Politics of Climate Change in Developing States Forest Lab

Political economy of the environment · Field lab

The Bupati's Cut

District heads in Indonesia sell logging permits into a shared wood market. Split a district or share out oil money, and watch how much forest gets cut.

Built on Burgess, R., Hansen, M., Olken, B. A., Potapov, P. & Sieber, S. “The Political Economy of Deforestation in the Tropics.” Quarterly Journal of Economics 127(4), 1707–1754 (2012).

Scenario
—
Forest Estate zone
Production: selective logging, by quota
Conversion: may be cleared for plantations
Protection: all logging illegal
Conservation: all logging illegal
Clearing
Cleared this year, within quota
Cleared this year, over quota (illegal)
Cleared in an earlier year
Not forest in 2001
Map marks
Oil & gas field
District border this year
1990 (Suharto-era) district border
Selected district
Why does this bupati cut this much?—
—
—

Where this year's cut comes from
The last permit sold, per 100 of wood price
Year—
2006
—
200120042008 · data ends2012
District splits—
Pick the year each new district is formed, or never. The defaults follow the real 2004–07 moratorium on new districts.
Levers
How likely Jakarta is to catch over-quota logging.
More oil money means a head has more to lose if caught.
35% is the paper's estimate (Table VII).
District
This district, 2001–2012ha / yr
Within quota Over quota Oil $/person ◆ close race◇ safe race

01The province over time

The gap between the solid line and each dashed line is how much that mechanism adds to or saves in clearing. Shaded years come after the paper's data ends.

Forest cleared per yearthousand ha

02The ledger

Every district in the selected year. Over quota is the illegal share of its permit cutting. Click a row to select it; click a heading to sort.

District Formed Cleared Over quota Market share Oil & gas $/cap Last margin Next election

03What the paper found

The simulation is tuned to these published estimates. The numbers here are the paper's, not the model's.

Headline estimates
+8.2%
More deforestation in a province, three years after it gains one district. The effect is as large in zones where all logging is illegal.
−3.4%
Lower wood prices for each extra district. More sellers, cheaper wood: the Cournot prediction.
−0.3%
Less logging per US$1 per person of oil and gas money, in the short run.
≈ 0
That oil effect three years later. It fades.
35%
Share of the oil effect that disappears after a district's first direct election for its head.
189 → 312
Districts on Indonesia's forest islands, 2000–2008.
The idea in three steps

1. District heads sell logging permits. Since 1999 the district forest office has answered to the bupati, and no log moves without its permit. A permit beyond the legal quota is a bribe.

2. More districts means more sellers. Logs can't be exported raw, so each province is its own wood market. Split a district and there is one more seller: more wood is cut and prices fall.

3. Oil money raises the stakes. A head caught logging illegally loses office. Oil money makes office worth more, so heads cut less. After direct elections, oil money seems to attract heads who take both kinds of rent.

04Your assignment

Work through these using the controls above. Cite specific districts, years and numbers from the ledger and the charts.

Questions
  1. One more seller. Choose “No new districts”, then “Paper baseline”, and compare the “From district splits” tile in 2008. Why does adding sellers raise total cutting even though each new district is smaller?
  2. Read the card. Select Kutai in 2001, then Kutai Kartanegara in 2004 after two splits. Which line of the first-order condition changed most, and why?
  3. Oil money. Try “Oil windfall ×2”. Which districts cut less? Why does Tana Tidung, with no wells of its own, get so much oil money per person?
  4. Elections. Move the post-election slider from 0% to 100% and watch Kutai Kartanegara after 2005. What kind of politician might an oil windfall attract?
  5. Policy. Compare fewer districts, tougher enforcement, and less oil money. Which saves the most forest by 2012, and which can Jakarta actually control?
How this model works

Each year every district head chooses how much forest to open to loggers. Selling more earns more bribes but lowers the price for everyone. Cutting past the legal quota risks detection, and detection costs more the more oil money a head stands to lose.

Oil money is shared by the real rule: half to the district with the field, half split among the rest. Direct elections start in 2005 and repeat every five years.

Each map cell is 2,500 ha. The model is tuned to the paper's estimates, but the map and forest are illustrative.