Changing preferences for environmental protection: evidence from volunteer behaviour

ABSTRACT

Over the past couple of decades, public awareness about environmental issues and concern for environmental protection appears to have increased substantially in Canada as it has in many other countries. This research has two objectives: first, to empirically assess the factors affecting voluntary participation in activities to protect the environment in Canada and to determine if participation has changed from 2000 to 2010, and second, to contemplate whether a change in environmental participation over time might be considered evidence of changing preferences, and thus demand, for environmental protection. Analysis makes use of data from Statistics Canada’s 2000, 2004, 2007 and 2010 Canadian Survey of Giving, Volunteering, and Participating used to estimate two IV probit models. Results suggest the likelihood of participating in environmental protection has increased over time providing a signal that preferences for environmental protection may have increased in Canada. Post-secondary education and social capital developed in youth have the largest impacts on the likelihood of participating. The results are expected to provide valuable information for public policy makers and environmental non-profit organizations.

KEYWORDS:

• Environmental economics
• volunteer
• public policy
• Canada
• environmental participation
• public goods

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. Public goods have the properties of non-rivalry and non-excludability. Non-rivalry states that any one person’s consumption of a public good, such as clean air, has no effect on the quantity of it available for others. Non-excludability states that it is impossible or very costly to exclude those who do not pay from consuming the good.

2. A probit selection model was initially estimated, but the Wald test of independent equations revealed that the decision to volunteer and the decision to participate in environmental activities are independent, thus the probit model was chosen.

3. Missing values consist of responses of ‘refused’ or ‘don’t know’.

4. Survey respondents less than 25 years of age were eliminated to restrict the analysis to those finished school.

5. Data for 2004 and 2007 CSVGP are also available, but are not included in the probit model due to the omission of some key variables measuring social capital.

6. The maximum likelihood approach is a method of estimating asymptotically efficient parameters of a statistical model.

7. The Blinder-Oaxaca decomposition for non-linear regression models (Sinning, Hahn, and Bauer 2008) is used to decompose the effect of the variable Year, representing time, on participation in environmental protection, and standard errors of the components are estimated using the bootstrap option. The results confirm the significant effect of Year on the dependent variable.

8. Experiments were conducted with interactions between Year and several of the socio-economic variables but the results were not statistically significant.

9. Many post-secondary institutions offer academic credit for service learning courses. For instance, Thompson Rivers University in British Columbia, Canada offers such courses within their bachelor of business administration degree, http://www.tru.ca/business/degrees/bba/dbe.html (accessed 15 May 2017).

10. For instance, in the city of Kamloops, British Columbia, the non-profit Volunteer Kamloops provides volunteer recruitment and referral services for any group involved in a volunteer project in the community, www.volunteerkamloops.org/about-volunteer-kamloops.htm.

11. For a discussion on crowding out by monetary incentives, see Ariely, Bracha, and Meier (2009).