Skip to main content

Climate Change and Recreation: Evidence from North American Cycling

Abstract

There is extensive research documenting the economic consequences of climate change, yet our understanding of climate impacts on nonmarket activities remains incomplete. Here, we investigate the effect of weather on recreation demand. Using data from 27 million bicycle trips in 16 North American cities, we estimate how outdoor recreation responds to daily weather fluctuations. We find empirically that cyclists dislike cold temperatures much more than hot temperatures, suggesting potential gains from warming. However, the overall response to extreme heat is mitigated, in part, by intraday adaptation towards recreating during cooler times of day. Combining these estimates with time-use survey data and climate projections, our models suggest annual surplus gains of $894 million from climate-induced cycling by mid-century.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10

Notes

  1. 1.

    https://www.bea.gov/outdoor-recreation/.

  2. 2.

    Here and throughout the paper, we use the term “leisure” interchangeably with warm-weather “outdoor recreation.”

  3. 3.

    Even so, we contend that our primary bikeshare estimates are much more precise than estimates based on time use surveys due to the unique nature of our data set and our estimand is more economically meaningful for projecting future leisure demand under climate change.

  4. 4.

    In complementary research, Leard and Roth (2019) estimate the welfare impact from fatal traffic accidents induced by climate change. They find large costs associated with traffic fatalities by the end of the century. Further, they posit that “voluntary exposure benefits” from more pedestrians, cyclists, and motorcyclists on the roads with higher temperatures offset the costs of fatalities, thus mitigating the consequences of climate change on traffic accidents. This latter finding is in line with our estimates for cycling, although our data permit us to estimate the benefits from climate-induced recreational activities.

  5. 5.

    Further, using automatic bicycle-traffic receptors, Nosal and Miranda-Moreno (2014) explore a similar dose–response relationship between weather and cycling along 14 routes in Montreal, Ottawa, Vancouver, and Portland, and Quebec over a span of 1-to-3 years. Their results are consistent with what we find, although we model our dose–response function more flexibly and include a wider variety of cities, over a longer timespan, with a broader set of controls and fixed effects, thus extending the validity of their results. In the present paper, we go several steps beyond Nosal and Miranda-Moreno (2014) to combine our estimates with climate projections and an economic welfare framework to predict climate change impacts on a national scale.

  6. 6.

    We observe more than 3500 years of cumulative time spent on bicycles in our full 8-year sample.

  7. 7.

    Note that the daily average temperature is the simple mean of the high and low temperature for that day. Therefore, a day in the \(>80\,^{\circ }\hbox {F}\) bin in our data may actually represent a very hot day with a high of \(95\,^{\circ }\hbox {F}\) and a low of \(70\,^{\circ }\hbox {F}\).

  8. 8.

    Formally, the temperature bins, in degrees Fahrenheit, are \((-\infty ,30]\), (30, 40], (40, 50], (50, 60], (70, 80], and \((80,\infty )\), with the (60, 70] bin omitted. The precipitation bins, in inches per day, are (0, 0.25], (0.25, 0.50], (0.50, 0.75], (0.75, 1], and \((1,\infty )\), with days with no precipitation omitted.

  9. 9.

    See Dell et al. (2014) and Carleton and Hsiang (2016) for broad reviews of this literature.

  10. 10.

    For more details on bikesharing programs, see Hamilton and Wichman (2018) and Wichman and Cunningham (2017).

  11. 11.

    Further, we interpolate seven observations total in Montreal, San Francisco, and Seattle for missing snowfall.

  12. 12.

    Because the primary purpose of this exercise is to establish robustness, we do not correct or impute missing values.

  13. 13.

    Graff Zivin and Neidell (2014) find that individuals substitute from outdoor leisure to indoor leisure at very high temperatures. Although the shape of our dose–response function would suggest a similar trend for bikeshare users, our coefficient on the highest temperature bin is statistically insignificant. This discrepancy may arise because we are studying different activities or because our top-coding of the highest temperature bin obscures the negative effect of more extreme temperatures. We discuss the latter in more detail below.

  14. 14.

    Because our outcome variables are natural logarithms and our variables of interest are dummy variables, we transform all marginal coefficients, as summarized by Wichman (2018), prior to interpreting our results as percentage effects. Specifically, percentage effects are calculated as

    $$\begin{aligned} \hat{g} = \exp \left( \hat{\beta } - 0.5 \hat{V}(\hat{\beta }) \right) - 1 \end{aligned}$$

    where \(\hat{\beta }\) is our estimated marginal effect and \(\hat{V}(\hat{\beta })\) is an estimate of its variance. Similarly, our measure of variance around percentage effects is calculated as

    $$\begin{aligned} \tilde{V}(\hat{g}) = \exp (2\hat{g}) \left[ \exp \left( -\hat{V}(\hat{\beta })\right) - \exp \left( -2\hat{V}(\hat{\beta })\right) \right] . \end{aligned}$$

    We present our primary results in log scale for comparison in Figure A.2 and in levels in Figure A.3.

  15. 15.

    We present these same figures with confidence intervals in Figure A.4. We omit confidence intervals in the main text for clarity.

  16. 16.

    Unfortunately, there are not enough observations in this range to further partition the \(>80^\circ\) bin. As described before, days with average temperatures in this range are actually very hot days, and they are relatively infrequent in our data, particularly for higher latitude cities (see Fig. 3).

  17. 17.

    See Tables A.3 and A.4 in the Online Appendix for city-specific regression results.

  18. 18.

    In a model of transportation demand, Cutter and Neidell (2009) categorize trips in an analogous manner. They consider trips during rush hours to be commuting (work-related), while classifying trips at other times of day as discretionary (leisure). They also comment that dividing their data into weekday and weekend samples would help sharpen the distinction between discretionary and commuting trips.

  19. 19.

    We have membership designations for all cities except Hoboken, Mexico City, and Pittsburgh.

  20. 20.

    We also find striking parallels between our cycling dose–response functions and those of Obradovich and Fowler (2017), who use survey data on a wide range of physical activity along with month-average weather observations; discrepancies in magnitude may arise because of measurement error from their use of coarser survey data.

  21. 21.

    Specifically, we construct our binned temperature variables with (1) average daily wet-bulb temperature, (2) maximum daily wet-bulb temperature, (3) average daily dry-bulb temperature, (4) maximum daily dry-bulb temperature, (5) maximum daily temperature in which each weather station within 100 km is weighted equally, (6) maximum daily temperature in which each weather station within 100 km is weighted by its inverse distance squared, (7) minimum daily temperature in which each weather station within 100 km is weighted equally, (8) minimum daily temperature in which each weather station within 100 km is weighted by its inverse distance squared, and (9) average daily temperature in which each weather station within 100 km is weighted equally. We derive the first four measures from the LCD weather data set, and the last five measures from the GHCN-Daily data set.

  22. 22.

    For context, the minimum temperature projection paired with the maximum precipitation projection (that is, roughly today’s temperature with rain nearly every day, a highly unlikely scenario), we see more than a 60% reduction in duration of trips. This projection aligns with our dose–response function for precipitation, where additional days with a large quantity of rain reduce demand by roughly the same quantity as this projection. Notably, our estimates are positive for the interquartile range, centered at a 4–6% increase, giving us confidence that climate change will tend to stimulate outdoor recreation demand.

  23. 23.

    For a small number of states, we did not observe positive time spent cycling, and we replaced those values with the national average.

  24. 24.

    Chan and Wichman (2018) analyze this welfare approximation framework, finding that it will tend to provide conservative estimates of welfare changes.

  25. 25.

    The Recreational Use Values Database (Oregon State University 2006) reports a mean consumer surplus value of $47.52 per day (2016 USD) for leisure cycling from 17 primary studies. Our analysis of the primary studies suggests an average of 2-h cycling trips, so we divide the given value by 2 to scale consumer surplus into an hourly measurement of $23.76 per hour.

  26. 26.

    The average state-level annual hourly wage used in our analysis is $27.02 (2015 USD), obtained from the US Census, Table H-8, Historical Income Tables.

  27. 27.

    This issue is discussed in greater detail by Chan and Wichman (2018). They also show that much of the value accrues via inframarginal recreation activity that becomes more pleasant, and these inframarginal gains in welfare will not be affected by substitution at the margin. Yet, if there are lingering doubts about this approach, we can interpret our welfare estimates as applicable only to cycling while conceding that there may be compensating losses from reductions in other leisure activities. Moreover, our analysis using guidance from the US Department of Transportation helps corroborate the scale of our estimated effects.

References

  1. Aguiar M, Hurst E, Karabarbounis L (2013) Time use during the great recession. Am Econ Rev 103(5):1664–1696

    Article  Google Scholar 

  2. Albouy D, Graf W, Kellogg R, Wolff H (2016) Climate amenities, climate change, and American quality of life. J Assoc Environ Resour Econ 3(1):205–246

    Google Scholar 

  3. Ashenfelter O, Greenstone M (2004) Using mandated speed limits to measure the value of a statistical life. J Polit Econ 112(S1):S226–S267

    Article  Google Scholar 

  4. Auffhammer M, Hsiang SM, Schlenker W, Sobel A (2013) Using weather data and climate model output in economic analyses of climate change. Rev Environ Econ Policy 7(2):181–198

    Article  Google Scholar 

  5. Barreca A, Clay K, Deschenes O, Greenstone M, Shapiro JS (2016) Adapting to climate change: the remarkable decline in the US temperature-mortality relationship over the twentieth century. J Polit Econ 124(1):105–159

    Article  Google Scholar 

  6. Becker GS (1965) A theory of the allocation of time. Econ J 75:493–517

    Article  Google Scholar 

  7. Bento AM, Roth K, Waxman A (2014) The value of urgency: evidence from congestion pricing experiments. Technical report, working paper

  8. Bertrand M, Duflo E, Mullainathan S (2004) How much should we trust differences-in-differences estimates? Q J Econ 119(1):249–275

    Article  Google Scholar 

  9. Burke M, Emerick K (2016) Adaptation to climate change: evidence from US agriculture. Am Econ J Econ Policy 8(3):106–140

    Article  Google Scholar 

  10. Burke M, Dykema J, Lobell DB, Miguel E, Satyanath S (2015) Incorporating climate uncertainty into estimates of climate change impacts. Rev Econ Stat 97(2):461–471

    Article  Google Scholar 

  11. Burke M, Hsiang SM, Miguel E (2015) Global non-linear effect of temperature on economic production. Nature 527(7577):235–239

    Article  Google Scholar 

  12. Cameron AC, Miller DL (2015) A practitioner’s guide to cluster-robust inference. J Hum Resour 50(2):317–372

    Article  Google Scholar 

  13. Cameron AC, Gelbach JB, Miller DL (2008) Bootstrap-based improvements for inference with clustered errors. Rev Econ Stat 90(3):414–427

    Article  Google Scholar 

  14. Carleton TA, Hsiang SM (2016) Social and economic impacts of climate. Science 353(6304):aad9837

    Article  Google Scholar 

  15. Chan NW, Wichman CJ (2018) Valuing nonmarket impacts of climate change: from reduced-form to welfare. Working paper 18–06, Resources for the Future

  16. Cutter WB, Neidell M (2009) J Environ Econ Manag 58(3):253–265

    Article  Google Scholar 

  17. Deacon RT, Sonstelie J (1985) Rationing by waiting and the value of time: results from a natural experiment. J Polit Econ 93(4):627–647

    Article  Google Scholar 

  18. Dell M, Jones BF, Olken BA (2012) Temperature shocks and economic growth: evidence from the last half century. Am Econ J Macroecon 4(3):66–95

    Article  Google Scholar 

  19. Dell M, Jones BF, Olken BA (2014) What do we learn from the weather? The new climate-economy literature. J Econ Lit 52(3):740–798

    Article  Google Scholar 

  20. Deschênes O, Greenstone M (2007) The economic impacts of climate change: evidence from agricultural output and random fluctuations in weather. Am Econ Rev 97(1):354–385

    Article  Google Scholar 

  21. Dundas SJ, von Haefen RH (2020) The effects of weather on recreational fishing demand and adaptation: implications for a changing climate. J Assoc Environ Resour Econ 7(2):209–242

    Google Scholar 

  22. Graff Zivin J, Neidell M (2014) Temperature and the allocation of time: implications for climate change. J Labor Econ 32(1):1–26

    Article  Google Scholar 

  23. Hamilton TL, Wichman CJ (2018) Bicycle infrastructure and traffic congestion: evidence from DC’s Capital Bikeshare. J Environ Econ Manag 87:72–93

    Article  Google Scholar 

  24. Hsiang S, Kopp R, Jina A, Rising J, Delgado M, Mohan S, Rasmussen DJ, Muir-Wood R, Wilson P, Oppenheimer M, Larsen K, Houser T (2017) Estimating economic damage from climate change in the United States. Science 356(6345):1362–1369

    Article  Google Scholar 

  25. Leard B, Roth K (2019) Voluntary exposure benefits and the costs of climate change. J Assoc Environ Resour Econ 6(1):151–185

    Google Scholar 

  26. Lechthaler F, Vinogradova A (2017) The climate challenge for agriculture and the value of climate services: application to coffee-farming in Peru. Eur Econ Rev 99:5–30

    Article  Google Scholar 

  27. Loomis J, Crespi J (1999) Estimated effects of climate change on selected outdoor recreation activities in the United States. In: Mendelsohn R, Neumann JE (eds) The impact of climate change on the United States Economy, chapter 11. Cambridge University Press, Cambridge, pp 289–314

    Chapter  Google Scholar 

  28. Mendelsohn R, Markowski M (1999) The impact of climate change on outdoor recreation. In: Mendelsohn R, Neumann JE (eds) The impact of climate change on the United States Economy, chapter 10. Cambridge University Press, Cambridge, pp 267–288

    Chapter  Google Scholar 

  29. Nosal T, Miranda-Moreno LF (2014) The effect of weather on the use of North American bicycle facilities: a multi-city analysis using automatic counts. Trans Res A Policy Pract 66:213–225

    Article  Google Scholar 

  30. Obradovich N, Fowler JH (2017) Climate change may alter human physical activity patterns. Nat Hum Behav 1:1–7

    Article  Google Scholar 

  31. Oregon State University. Recreational use values database, 2006. Data Accessed 4 Sept 2017. http://recvaluation.forestry.oregonstate.edu/

  32. Schlenker W, Roberts MJ (2009) Nonlinear temperature effects indicate severe damages to US crop yields under climate change. Proc Nat Acad Sci 106(37):15594–15598

    Article  Google Scholar 

  33. Sinha P, Caulkins ML, Cropper ML (2018) Household location decisions and the value of climate amenities. J Environ Econ Manag 92:608–637

    Article  Google Scholar 

  34. Sun L, Kunkel KE, Stevens LE, Buddenberg A, Dobson JG, Easterling DR (2015) Regional surface climate conditions in CMIP3 and CMIP5 for the United States: differences, similarities, and implications for the U.S. National Climate Assessment. NOAA technical report NESDIS 144, p 111

  35. Tol RSJ (2009) The economic effects of climate change. J Econ Perspect 23(2):29–51

    Article  Google Scholar 

  36. U.S. Department of Transportation (2014) The value of travel time savings: Department guidance for conducting economic evaluations. Technical report, Office of the Secretary of Transportation

  37. USDA (2012) USDA plant hardiness zone map. http://planthardiness.ars.usda.gov. Accessed 8 Sept 2017

  38. Walther G-R, Post E, Convey P, Menzel A, Parmesan C, Beebee TJC, Fromentin J-M, Hoegh-Guldberg O, Bairlein F (2002) Ecological responses to recent climate change. Nature 416(6879):389–395

    Article  Google Scholar 

  39. Wichman CJ (2018) Interpreting non-linear semi-elasticities in reduced-form climate damage estimation. Clim Change 148(4):641–648

    Article  Google Scholar 

  40. Wichman Casey J, Cunningham B (2017) Notching for free: do cyclists reveal the value of time? Working paper 17–17, Resources for the Future

  41. Wolff H (2014) Value of time: speeding behavior and gasoline prices. J Environ Econ Manag 67(1):71–88

    Article  Google Scholar 

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to Casey J. Wichman.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

We are grateful to Margaret Walls, Dave Keiser, Mirco Dinelli, Christine Crago, and Arik Levinson for helpful comments and suggestions. We also thank seminar and conference participants at the University of Minnesota, Cornell University, Colgate University, Bates College, Yale University, the Southern Economic Association Annual Meetings, the Heartland Environmental and Resource Economics Workshop at the University of Illinois, and the Northeast Workshop on Energy Policy and Environmental Economics at Columbia University for useful feedback. DaWon Lee provided exceptional research assistance.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (pdf 809 KB)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Chan, N.W., Wichman, C.J. Climate Change and Recreation: Evidence from North American Cycling. Environ Resource Econ 76, 119–151 (2020). https://doi.org/10.1007/s10640-020-00420-5

Download citation

Keywords

  • Leisure demand
  • Outdoor recreation
  • Climate change
  • Adaptation
  • Nonmarket damages
  • Time allocation

JEL Classification

  • J22
  • Q54
  • R49