Measuring Environmental Benefits: Revealed Preference Approaches

1. Measuring Environmental Benefits: Revealed Preference Approaches Travel Cost and Hedonic Methods

2. Motivation: Generic Group Projects • Chief “smokey” at Yosemite National Park needs to raise $10 million for bus system in park through increase in entrance fee. What should be the fee increase to pay for this? • You have been contracted to examine cutting back flights at SBA because of noise in surrounding area. Question: Is noise a problem that justifies this? How much does noise depress property values?

3. Revealed preference approaches (two most common) • Travel Cost Model: use data from actual visitations, estimate cost of travel, derive demand curve for visits to the “site”. (A type of “Household Production” model) • Hedonic Price Method: compare products with similar attributes but one “bundles” an environmental good, derive demand for the environmental good. • House prices influenced by environmental amenity (eg noise) • Wages influenced by riskiness of job

4. 1. Travel Cost Model: Yosemite • Need to know the demand for park visits (note: this reflects use value only) • Current entrance fee=$20. (Is this related in any way to the park’s value? How?) • Goal: empirically develop demand curve for visits • Typical visitor • L = # hours worked by person at wage w. • P0 = out-of-pocket expenses to visit Yosemite, • F = entrance fee. • t = travel time, s = visit time • Price of a trip: p = [P0+w(t+s) + F]

5. Effective price of trip Price of a trip: p = [P0+w(t+s) + F] • Notice opportunity cost of time (w) • This assumes we value travel time and visitation time at the wage rate of the individual. • Value of time ranges, but is often estimated at 1/3 or 1/2 the wage rate.

6. Objective • Want to derive a demand curve for visits to Yosemite. • What can we do with a demand curve? • Calculate consumer surplus (benefits) – review concept on board… • Can calculate use-value of Yosemite • Can determine cost to consumers from e.g. entrance fee (from $20 to F1)

7. Demand for visits to Yosemite $ . NEW Consumer Surplus . . . . . OLD Consumer Surplus . . . . . . F1 . . . F0 Demand . V1 Visits V0

8. Procedure [1 of 3] 1. Station students at park entrance on several “random” days. • Ask visitors (1) zip code, (2) other stuff (mode of travel, $ spent, socioeconomic characteristics…) • Scale up answers to entire year, over entire pop: • # visits/zip code/year to park • Use knowledge of total number of visits to park per yr 2. Calculate travel cost from each zip code • Use travel time, travel costs, wages in zip code • This, with the entrance fee, is the “price” of a visit: p = TC +F 3. Sort zip codes into “zones” of equal travel cost • E.g. Sacramento, Santa Barbara, Germany, … , etc.

9. Travel cost “zones” Yosemite Z=1 Z=4 Z=3 Z=2 Zones have equal travel cost within each zone.

10. Procedure [2 of 3] 4. For each zone, • Calculate population (Pz) of zone • Estimate number of visits (Sz) from zone • Calculate visitation rate: vz = Sz/Pz. 5. Estimate relationship between price and visitation rate • v = f(π,y,d) = f(TC+F,y,d) • Plot price (pz) vs. visitation rate (vz) – scatter plot • Perform multiple regression to control for income (y) and other variables (d) Price, πz . . f . . . . . . . vz

11. Procedure (Step 3 of 3) 6. Vary F from 0 to some upper bound; for each F: • Calculate visits for each zone, using v=f(TC+F,y,z) and characteristics of zone • Add up over all the zones to obtain total visits to the park for each F 7. Voila: Demand curve! • Visits as a function of entrance fee, F

12. Demand curve Demand for visits to Yosemite Entrance Fee, F Number of visits, V

13. Finish your analysis • Use demand curve to advise head Smokey • Calculate revenue from different park entrance fees

14. 2. Hedonic pricing to value risks • Do you trade off risks to your life with money?

15. 2. Hedonic pricing to value risks • Do you trade off risks to your life with money? • Observe: workers willing to undertake risk for increased pay • Observes wage-risk tradeoffs in labor market • Hedonics: Compare different occupations with different risks of mortality • Assumes workers are aware of risks and that they are perfectly internalized. • Assumes only real difference between occupations is level or risk, but can control for some other variables.

16. Occupations similar except risks

17. VSL: Willingness to pay for marginal reduction in risk to life Wage Calculate W(p). dW/dp = VSL (Value of Statistical Life) Wrong interpretation: change in wage when risk changes from p = 0 to π = 1.0. VSL typically $3-$6 million From wage-risk studies . VSL . . . . . . Prob death (p) … 1.0 .0001 .0002

18. “Correct Interpretation” • Suppose a new health regulation decreases risk of death by .1% for all citizens in a city with population 100,000. • Expected number of “statistical lives” saved by regulation is .001*100,000=100. • If VSL=$6M, then benefits are $600M. • Correct interpretation is that each member of population benefits by .001*6M=$6,000. Since 100,000 people benefit, total benefit is $600M.

19. VSL Studies (1990 US$) • Australia (1984): $3.3 million • Japan (1986): $7.6 million • US (1982): $16.2 million • Canada (1979): $3.6 million • UK (1977): $2.8 million • US (1976): $6.5 million Caution: ignores age & health

20. Hedonic Analysis of Property Values Hedonic Price Function House Prices x x x x x x x x x x x x Noise Level

21. What is missing? • All other factors that affect house price • Lot size • Rooms/Bathrooms • House age • School quality • Etc. • Can “control” for these using multiple regression….acts like an experiment.

22. Hedonic Price Analysis • Estimate marginal effect of noise of house prices • Compute price effects of reduced noise at airport • Approximation of willingness-to-pay for noise reduction

23. Conclusions • Revealed preference methods desirable for valuing environmental benefits • Relies on “fortuitous” association of markets with environmental goods – not that common • Two basic methods • Travel cost (household production) • Hedonic (typically housing or wage-risk studies)

24. Valuing recreational non-consumptive use of MPAs • Bren GP 2004 • What is the value of MPAs for recreational non-consumptive users? • Counterfactual is critical • Reserves had just been implemented • Zonal travel cost on charter boats • Condition on: socioeconomic, expected quality, expected impact of reserves, knowledge of reserves • Regress VR on travel cost, income, substitutes, etc. • Result: value of recreation was high (around $10-$15 surplus per visit), but current value of MPAs low

25. Demand with and without MPAs

26. Predicted Visitation

27. Average Visitor Surplus