In Part 1 I reviewed the scant U.S. literature on outdoor recreation and climate change and in Part 2 I looked at the effects of temperature and precipitation on recreation participation in three sectors: boating, freshwater fishing and hunting. In Part 2 using data from the National Survey on Recreation and the Environment, the main conclusion is that:
In contrast to previous studies, I'm finding that climate change decreases boating and hunting participation with no effect on freshwater fishing. If I was really serious about this I'd need to dig further but it does bring into question the aggregate data national estimates from a few years ago (Part 1).
In part 3 I'm digging deeper into the freshwater fishing sector, with different data, to see if my result will hold up since other studies find an improvement in the economic value of freshwater fishing with a warmer and wetter U.S.
The bottom line is that climate change from 2001 to 2098 will result in an annual loss of welfare of about $1.13 billion by 2098, 9% of the total economic value of freshwater recreational fishing. With a 2% discount rate, the present value of losses in freshwater fishing economic value is $16.6 billion. With a 7% discount rate the present value of losses is $2.33 billion.
In part 2 I said:
I could get days participation from the National Survey of Fishing, Hunting and Wildlife-Associated Recreation but that's a lot of work and this stupid blog is not a peer-reviewed journal.
I guess this blog isn't all that stupid because I pulled out my 2001 FHWAR CD and got the freshwater fishing data in great shape: cleaned, scaled and weighted (pun intended). There are over 6751 day trip anglers in the final data.
The gain from using the FHWAR data is that you can estimate recreation participation and then days participation in a two stage model. Using the same methods as in Part 2 I've estimated models for 14 single species and species groups. The target probability and average days fished for those who target are in parentheses (note the probabilities add to greater than one since each fishing day may have multiple targets):
- Crappie (21%, 7)
- Panfish (25%, 8)
- Striped bass (14%, 7)
- Black bass (27%, 21)
- Catfish (24%, 8)
- Walleye (11%, 9)
- Sauger (<1%, 4)
- Northern pike (6%, 7)
- Trout (29%, 8)
- Salmon (4%, 6)
- Steelhead (<1%, 4)
- Other1 (4%, 6)
- Other2( 1%, 5)
- Anything (13%, 4)
I used linear models for both participation and days fished (I'm saving the 2-stage probit/truncated [i.e., Cragg] model for the JPE). There are fourteen models so I won't present all ... but the trout model, the species which suffers the greatest losses, is pictured to the right (click on the thumbnail for a larger image).
For trout fishing, increases in January temperatures decreases participation and the number of days. Increases in April temperatures does the opposite. Increases in July temperatures reduces the number of days. Increases in October temperatures increases participation. Increases in January precipitation decreases participation. Increases in April and July precipitation increases participation. Increases in July and October precipitation decreases days.
For each model I next predicted the baseline participation rate, simulated the participation rate with climate change in 2098 and computed the difference. I then did the same thing with days fished. Next, I scaled the number of participants and days from the sample so that nothing was double-counted when aggregated (see page 4 of the FHWAR report [PDF]). When scaled, my national model predicts that 34.1 million anglers will fish on 567 million days.
The model predicts the following change in participation rates and days fished (for the subsample that targeted that specie(s)):
- Crappie (7%, -1)
- Panfish (2%, -1)
- Striped bass (4%, -1))
- Black bass (2%%, -3)
- Catfish (12%, -1)
- Walleye (2%, 2)
- Sauger (0%, -5)
- Northern pike (-1%, -1))
- Trout (-18%, -3)
- Salmon (-2%, -1)
- Steelhead (<1%, 2)
- Other1 (-0%, -3)
- Other2( -2%, -0)
- Anything (3%, -1)
Target participation is expected to increase in 8 out of 14 species. Days fished is expected to decrease for 12 out of 14 species. Added up, the model predicts a total of 517 million fishing days with 2098 temperature and precipitation, a loss of almost 9% of days from the 2001 baseline.
Attaching consumer surplus values ($22.78 in 2001 dollars) obtained from Table 1 of Rosenberger and Loomis [PDF] to the lost fishing days yields a lost annual consumer surplus value of $1.13 billion in 2098.
The aggregate loss over 98 years can be estimated by assuming no loss in 2001 and a constant increase in losses each year out to 2098. With a 2% discount rate, the present value of losses in freshwater fishing economic value is $16.6 billion. With a 7% discount rate the present value of losses is $2.33 billion.
Whew!
