Skip to main content

Does Absolution Promote Sin? A Conservationist’s Dilemma


This paper shows that households signing up for a green program exhibit an intriguing behavioral rebound effect: a promise to fully offset customers’ carbon emissions resulting from electricity usage increases their energy use post-adoption by 1–3%. The response is robust across empirical specifications, and is consistent with an economic model of rational energy consumption. Our results provide a cautionary tale for designing green product strategies in which the adoption of a product may lead to unexpected consequences.

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

Fig. 1
Fig. 2
Fig. 3
Fig. 4


  1. An extensive literature exists on the so-called “rebound effect”, which broadly addresses the potential general equilibrium effects of energy efficiency and other conservation policies (Borenstein 2015; Gillingham et al. 2016). Additionally, recent psychological literature has devoted a substantial effort to document the presence of moral licensing through a variety of laboratory and small scale field experiments (Effron and Monin (2010); Kouchaki (2011); Merritt et al. (2010)). Since consumer choices, especially in the environmental arena reflect social and moral values, this appears to be one promising mechanism that could explain the psychology behind the adoption of carbon offsets and subsequent change in energy consumption (Jacobsen (2010); Kotchen (2009)). Green markets also affect social welfare, underlying the importance of understanding consumer choice in this setting (Kotchen 2006).

  2. A particularly poignant story of fraud in the market for “environmental indulgences” was reported by the Christian Science Monitor in April 2010, which revealed how the Vatican was convinced to purchase carbon offsets that would have lead to the Vatican becoming the first carbon free state, but which were never implemented. The purchased offsets were meant to be used for the planting of millions of trees in Hungary. As it turned out the trees were never planted and the Hungarian company abruptly closed down at the end of 2007. See

  3. While the experimental literature seems to find that behavioral factors substantially increase the adoption of green power programs, we should caution that the effect may not be universally effective. In an unpublished experiment, one of the authors worked with a major utility and sent 50,000 letters encouraging utility customers to sign up for a green power program using several well-documented behavioral nudges such as social pressure. The experiment did not generate a single adoption but did lead to three complaint letters being sent to the utility and the experiment was not published. This might indicate that publication bias also plays a role in the claimed successes of nudges.

  4. The stock of greenhouse gases is large enough that any individual’s contribution is infinitesimal. We nonetheless include it in order to keep the model applicable to a broad class of closely-related environmental considerations (e.g. effects on local criteria pollutants).

  5. We interpret the parameter \(\delta \) as a behavioral parameter but remain agnostic about the precise behavioral/psychological mechanism influencing it. It is possible that \(\delta \) reflects the degree of understanding or awareness of the social cost of pollution which is a function of education, social, religious and political beliefs. Similarly, since a large share of the cost is likely to be incurred in the future, variation in the \(\delta \) parameter may reflect individual inter-temporal discount rates and concern for future generations.

  6. For the purpose of our stylized model we ignore non-linear electricity pricing, which is common in the residential electricity market, but does not affect the results of interest.

  7. Note that as a consequence of our stylized model we are ignoring income effects. Basic economic intuitions tell us that, all else equal, higher income households are more likely to sign-up for the program. As we shall see later in the paper this is indeed the case. It is easy to derive the income effect in the current framework under a suitable reformulation of the utility function. This however does not change the other implications of the model and we abstract from the income effect in favor of notational simplicity.

  8. The baseline price of electricity is determined by usage with five tiers ranging between $0.12233 and $0.34180 per kWh.

  9. The price is well-within the range of per ton prices usually encountered for carbon offsets in the offset markets, but substantially lower than common estimates of the social cost of carbon which typically is measured in the $30–50 per ton range, depending on the chosen discount factor.

  10. Climate Action Reserve describes their protocol in detail in a program manual that is publicly available:

  11. Source: International Carbon Bank and Exchange, California Energy Commission.

  12. Further details on the CS program are available through the detailed annual reports issued by PG&E. The figures quoted are from the report for 2010 which is publicly available at:

  13. This rate is not time-of-use dependent. Throughout this analysis we do not consider low-income households. These households are on special variants of this electric rate, and PG&E took extra steps to ensure that they were aware of being in the program. The program was available to all customers on an opt-in basis but low-income customers on special rates were subsequently de-enrolled by the utility.

  14. It is quite common for anomalies to appear in any electricity billing dataset. For example, billing errors may occur that create unrealistic patterns (say, zero dollars in June but twice the seasonally-adjusted expectation for July). We remove these months from the data, but see no justification for dropping the entire household time-series due to such brief idiosyncratic billing events. As such, we allow for a small number of absent months from our pseudo-balanced panel. In any case, results are robust to changes in these minor sampling assumptions.

  15. Data on individual households was purchased by the utility as part of the regular business processes. Given the availability of complete address information in their administrative database we do not think selection to be a major factor.

  16. For example, San Francisco is in zone 3 while zone 12 corresponds to the Northern Central Valley which experiences substantially hotter summers than zone 3.

  17. One of the authors of this paper purchased data on himself from the same provider and the purchased information found the information to be accurate, although it slightly underestimated his environmentalist credentials.

  18. We have also used month and year dummies to control for seasonality, which impose a lower computational burden as that specification requires fewer parameters. The results were almost identical. Later we shall discuss more robustness checks to account for seasonality in our specifications.

  19. Further, from a program evaluation perspective, our setting is also appropriate. Under no realistic implementation of this sort of program would the allocation of offsets be random. The same is not true for allocation of permits under a cap-and-trade regime, a very different setting than what we study here.

  20. Note that effects in this direction would imply an upward sloping demand curve if price were the driving factor.

  21. Jacobson et al. (1992).

  22. The results are not affected by the choice of the Weibull model. More complicated duration models do not change the results.

  23. A specification that included household by month-of-year controls produced results that were essentially noise. However, such an over-specified model does not leave adequate residual variation to credibly identify an effect.



  • Ayal S, Gino F (2011) Honest rationales for dishonest behavior. Exploring the causes of good and evil, chap. Honest rationales for dishonest behavior. APA

  • Baker JS, McCarl BA, Murray BC, Rose SK, Alig RJ, Adams D, Latta G,Beach R, Daigneault A (2010) Net farm income and land use under a USgreenhouse gas cap and trade. In: Agricultural and Applied EconomicsAssociation. Policy Issues 7: April 2010. p 5

  • Bloomberg (2012) State of the voluntary carbon market 2012. New energy finance, ecosystem marketplace

  • Bolderdijk JW, Steg L, Geller ES, Lehman P, Postmes T (2013) Comparing the effectiveness of monetary versus moral motives in environmental campaigning. Nat Clim Chang 3(4):413

    Article  Google Scholar 

  • Borenstein S (2015) A microeconomic framework for evaluating energy efficiency rebound and some implications. Energy J 36:1–21

    Google Scholar 

  • Brown TR, Elobeid AE, Dumortier JR, Hayes DJ (2010) Market impact of domestic offset programs

  • Clayton S, Devine-Wright P, Stern PC, Whitmarsh L, Carrico A, Steg L, Swim J, Bonnes M (2015) Psychological research and global climate change. Nat Clim Chang 5(7):640

    Article  Google Scholar 

  • Costa D, Kahn M (2013) Energy conservation “nudges” and environmentalist ideology: evidence from a randomized residential electricity field experiment. J Eur Econ Assoc 11(3):680–702

    Article  Google Scholar 

  • DellaVigna S, List J, Malmendier U (2012) Testing for Altruism and social pressure in charitable giving. Q J Econ 127(1):1–56

    Article  Google Scholar 

  • Ebeling F, Lotz S (2015) Domestic uptake of green energy promoted by opt-out tariffs. Nat Clim Chang 5(9):868–871

    Article  Google Scholar 

  • Effron D, Monin B (2010) Letting people off the hook: When do good deeds excuse transgressions? Pers Soc Psychol Bull 36:1618–1634

    Article  Google Scholar 

  • Gillingham K, Rapson D, Wagner G (2016) The rebound effect and energy efficiency policy. Rev Environ Econ Policy 10(1):68–88

    Article  Google Scholar 

  • Gneezy U, Rustichini A (2000) A fine is a price. J Legal Stud 29:1–18

    Article  Google Scholar 

  • González-Ramírez J, Kling CL, Valcu A (2012) An overview of carbon offsets from agriculture. Annu Rev Resour Econ 4(1):145–160

    Article  Google Scholar 

  • Graff-Zivin J, Lipper L (2008) Poverty, risk, and the supply of soil carbon sequestration. Environ Dev Econ 13(3):353–373

    Article  Google Scholar 

  • Harding M, Lamarche C (2011) Least squares estimation of a panel data model with multifactor error structure and endogenous covariates. Econ Lett 111(3):192–199

    Article  Google Scholar 

  • Herberich D, List J, Price M (2011) How many economists does it take to change a light bulb? A natural field experiment on technology adoption, Working Paper

  • Hodrick R, Prescott E (1997) Post war business cycles: an empirical investigation. J Money Credit Bank 29:1–16

    Article  Google Scholar 

  • Hsieh D, Manski C, McFadden D (1985) Estimation of response probabilities from augmented retrospective observations. J Am Stat Assoc 80:651–662

    Article  Google Scholar 

  • Jacobsen G (2010) Do environmental offsets increase demand for dirty goods? Evidence from residential electricity demand

  • Jacobsen G (2011) The Al Gore effect: an inconvenient: truth and voluntary carbon offsets. J Environ Econ Manage 61:67–78

    Article  Google Scholar 

  • Jacobsen GD, Kotchen MJ, Vandenbergh MP (2012) The behavioral response to voluntary provision of an environmental public good: evidence from residential electricity demand. Eur Econ Rev 56(5):946–960

    Article  Google Scholar 

  • Jacobson L, Lalonde R, Sullivan D (1992) Earnings losses of displaced workers. Am Econ Rev 83:685–709

    Google Scholar 

  • Kotchen M (2006) Green markets and private provision of public goods. J Polit Econ 114:816–834

    Article  Google Scholar 

  • Kotchen M (2009) Voluntary provision of public goods for bads: a theory of environmental offsets. Econ J 119:883–899

    Article  Google Scholar 

  • Kotchen M, Moore M (2007) Private provision of environmental public goods: household participation in green-electricity programs. J Environ Econ Manage 53:1–16

    Article  Google Scholar 

  • Kouchaki M (2011) Vicarious moral licensing: the influence of others’ past moral actions of moral behavior. J Person Soc Pychol 101:702–715

    Article  Google Scholar 

  • Litvine D, Wüstenhagen R (2011) Helping “ light green” consumers walk the talk: results of a behavioural intervention survey in the Swiss electricity market. Ecol Econ 70(3):462–474

    Article  Google Scholar 

  • Manski C, Lerman S (1977) The estimation of choice probabilities from choice based samples. Econometrica 45:1977–1988

    Article  Google Scholar 

  • Merritt A, Effron D, Monin B (2010) Moral self-licensing: when being good frees us to be bad. Soc Personal Psychol Compass 4(5):344–357

    Article  Google Scholar 

  • Monin B, Miller D (2001) Moral credentials and the expression of prejudice. J Personal Soc Psychol 81:33–43

    Article  Google Scholar 

  • Pesaran H (2006) Estimation and inference in large heterogeneous panels with a multifactor error structure. Econometrica 74(4):967–1012

    Article  Google Scholar 

  • Pichert D, Katsikopoulos KV (2008) Green defaults: information presentation and pro-environmental behaviour. J Environ Psychol 28(1):63–73

    Article  Google Scholar 

  • Ravn M, Uhlig H (2002) On adjusting the Hodrick–Prescott filter for the frequency of observations. Rev Econ Stat 84(2):371–376

    Article  Google Scholar 

Download references


Funding was provided by Stanford Precourt Institute for Energy (US).

Author information

Authors and Affiliations


Corresponding author

Correspondence to David Rapson.

Additional information

Publisher's Note

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

We are grateful to PG&E for sharing the data and for assisting with the data preparation. We thank Marcel Priebsch for excellent research assistance. We are grateful to Hunt Allcott, Antonio Bento, Wesley Hartmann, Grant Jacobsen, Matthew Kotchen, Prasad Nair, Ted O’Donoghue, Olivier Rubel, Rob Stavins and seminar participants at Cornell, Dartmouth, Harvard, Stanford, UC Berkeley and UC Santa Barbara for excellent comments. The Precourt Energy Efficiency Center at Stanford University generously funded this project.

A Appendix

A Appendix

A.1 Background on the Voluntary Carbon Offset Market

By definition one carbon offset corresponds to the removal or neutralization of one metric ton of \(\text {CO}_2\) or an equivalent amount of other gases such as methane, nitrous oxide, hydrofluorocarbons, perfluorocarbons or sulphur hexafluoride, all of which contribute to the greenhouse effect.

The basic intuition behind this market is that since carbon emissions contribute to a global stock, abatement in one part of the world is equivalent to abatement elsewhere. If marginal abatement costs differ across regions, it should be possible to offset one’s emissions in an indirect, cost effective way. Each carbon offset is generated as a result of a specific environmental project, most of which can be located at a considerable distance from the buyer of the carbon offset. The range of environmental projects which can offset carbon emissions is vast and range from clean energy generation such as wind power to forest conservation to livestock waste management. Companies engaging in these environmental projects can claim to have produced offsets as long as the carbon removed is in excess of what would have been occurred in the absence of the offset. For example, one cannot label a project as generating carbon offsets if it would have happened without the funding generated by carbon offsets. The most popular types of projects involve either agricultural land use or forestry and many are located in developing countries.

In order to guarantee the validity of the carbon offsetting claims, common practice in the industry is to gain third-party certification for projects. In some cases buyers transact directly with the offset generator, while in other cases sophisticated markets have developed which allow carbon offsets to be traded. At the present, numerous companies are involved in marketing and trading carbon offsets, and numerous concerns about permit legitimacy underscore the importance of the verification process. It is often very difficult to accurately verify an alleged certification, since offsets are often purchased for future projects and the baseline level of emissions is often debatable. To address these concerns, PG&E engaged a reputable third-party certification organization, Climate Action Reserve,Footnote 24 to verify the emissions reductions associated with its offsets. PG&E circus a broad solicitation to purchase offsets, and each ton of greenhouse gas abated must then be certified by the Climate Action Reserve. Nonetheless, the implications of the findings in this study should be considered in the context of the broader market.

While a majority of the carbon market consists of regulated markets such as the EU Emission Trading Scheme, which covers the emissions of several thousand energy intensive European companies, a relatively small fraction of the market consists of voluntary carbon offsets. In 2011 the voluntary carbon market had a total value of $576 million down from a peak of $728 million in 2008 (Bloomberg 2012). In spite of the decline in this market resulting from the recent financial crisis, the voluntary offset market is particularly popular in North America where offsets are purchased through numerous over-the-counter contracts. Since American buyers appear to prefer more local projects, the majority of carbon offsetting projects red to the voluntary market are now to be found within the US.

Individual consumers are currently offered a number of ways in which they can offset their carbon footprint. Airline passengers are routinely asked if they wish to offset their carbon footprint resulting from air travel (sometimes at considerable cost). Voluntary carbon offsetting programs have also recently been offered to residential consumers in order to offset the carbon footprint resulting from everyday energy use at home.

A.2 Seasonality Corrections

A.2.1 Interactive Effects

One concern is that the seasonal controls (aggregate month-by-year effects) may be insufficiently capturing important sources of heterogeneity in time-varying unobservables. For example, if young households are more vulnerable to economic shocks or adverse weather conditions, electricity usage patterns within this cohort may exhibit higher volatility which is not fully captured by the mean time effects we control for. This would represent an unobserved source of heterogeneity which may potentially bias our estimation. At the core of this problem is the extent to which the assumption on the error term in Eq. 5 is correct. We can think of the total error term as given by:

$$\begin{aligned} u_{it}=\alpha _t+\gamma _i+\epsilon _{it}, \end{aligned}$$

where \(\epsilon _{it}\) is iid across households and time.

We are concerned that the true model may in fact have interactive effects of the form:

$$\begin{aligned} u_{it}=\gamma _i+\omega _iF_t+\epsilon _{it}. \end{aligned}$$

where \(F_t\) is an aggregate effect that is scaled by some demographic variable, \(\omega _i\). Such a model is not directly estimable, but the presence of demographic variables offers a way to test the validity of this concern and, at least partially, to eliminate it. We can consider various proxies for \(\omega _i\) without necessarily requiring the unobserved trends \(F_t\) to vary over each individual. We can divide the sample into different groups (e.g. by age quartiles) an estimate a model under the following assumption:

$$\begin{aligned} u_{it}=\gamma _i+\omega (1(i \in Group ~ g))F_t+\epsilon _{it}, \end{aligned}$$

where \(\omega \) is now observable for each i and \(F_t\) is approximated by year-month specific indicator variables.

In Table 6, we present the coefficients from Eqs. 3 and 4, but with month and years controls interacted with an array of demographic characteristics, and each cell again represents the estimate of \(\beta \) from a separate regression. The results are consistent with the baseline difference-in-differences (zero effect) and first differences estimates (1.5–2.5% rebound).

Table 6 Robustness of model specifications to interactive trends

It also possible that the true model has more than one interactive effect and may be multidimensional with an error term that has an unknown nt factor structure. As an additional robustness check we have implemented the control variable approach described in Pesaran (2006) and Harding and Lamarche (2011). The result was also very similar to the baseline specification indicating that our estimation is robust to a variety of interactive effects specifications.

A.2.2 Time Series Filtering

Finally, it is possible that the interactive effects specifications discussed above may not capture all the relevant heterogeneity. One such case is when the error term can be decomposed into two components, \(\epsilon _{it}\) which is iid across observations and another component \(v_{it}\) which is non-stationary and exhibits seasonality and trending behavior. This case requires us to filter the time series of electricity consumption for each household. We use the H–P (Hodrick–Prescott filter) commonly used in macroeconomics. Consider some arbitrary household. To remove the trend from log electricity consumption for that household \(ln(k_t)\), we decompose \(ln(k_t) = \tau _t + c_t\), where \(\tau _t\) is the trend, and \(c_t\) is the cyclical component. \(\tau _t\) is estimated by solving

$$\begin{aligned} \min _{\{\tau _t\}} \, \sum _{t=1}^T (x_t -\tau _t)^2 + \mu \sum _{t=2}^{T-1} [(\tau _{t+1}-\tau _t)-(\tau _t-\tau _{t-1})]^2 \end{aligned}$$

The parameter \(\mu \) penalizes variation in the first difference of the trend and is set to 6.25, 1600, and 129, 600 (Hodrick and Prescott 1997; Ravn and Uhlig 2002). We then use the filtered residual \(c_t = ln(k)_t - \tau _t\) as the new dependent variable in Eq. 5. Figure 5 shows the intuition behind the H–P filter for the time series of consumption for one arbitrary household. Notice the effect of the choice of the smoothing parameter \(\mu \). This suggests that a small value of \(\mu \) is required in order to remove the cyclical component. It is important to note an important limitation of this approach. Since this method is applied at the individual level it will remove to a large degree any pre and post adoption trends in behavior (such as pre-adoption conservation and the disappearance of the “rebound effect” post-adoption. The individual filtering approach however allows us to focus on the immediate post-adoption period and remains reliable in detecting an immediate jump in consumption after the household enrolls in the CS program.

Fig. 5
figure 5

Example of different smoothing parameters in the HP filter

Fig. 6
figure 6

Estimation of the dynamic effect of adoption in event time after filtering individual observations

In Fig. 6 we present the results for the event study conducted on residuals from different applications of the H–P filter with various degrees of smoothing \(\mu \). We see that even though the H–P filter removes some of the pre and post adoption trends it continues to clearly identify a discontinuity in usage at the time of adoption. The effect is slightly diminished but we still observe a 2% increase in consumption post adoption.

A.3 Profile of Adopters

Given the surprising and robust behavioral response resulting from adoption, it is important to understand if customers may differ along predictable demographic dimensions. This has substantial managerial implications for program design and customer targeting. In order to investigate how selection into adoption is driven by observable characteristics of the households, we construct a household specific variable \(T_{i}\), which equals one if household i signed up for the CS program, and zero otherwise. We use households that never sign up (\(T_{i}=0\)) as our reference group and we formally identify the model by normalizing the corresponding coefficients. We wish to model the probability of adopting the CS program by household i conditional on observed covariates \(x_i\), \(Pr(T_{i}=1|x_i)\). We encounter one important technical limitation however. By construction, our sample is non-random. In fact our very data request from PG&E was not formulated as a random sample of all PG&E customers. Given the low number of CS program adopters relative to PG&E’s large residential customer base this would have been impractical as adoption would have been a very rare event in a random sample. As such we chose to sample conditional on adoption status. The final sample contains a sizable proportion of CS program adopters and a very small but representative sample of the population of non-adopters subject to the restrictions on residence imposed earlier to achieve balance.

This sampling framework is usually referred to as “choice based sampling” or “retrospective sampling” since it uses the ex-post outcomes as part of the sampling frame. It is well known in this setting that estimation by maximum likelihood leads to inconsistent parameter estimates (Manski and Lerman 1977). While several approaches are available to address this issue, consistent estimates are typically obtained by pseudo-maximum likelihood where observations are weighted by a factor \(\mu _j=n_j/(NPr(T_j))\), where \(n_j\) corresponds to the observed sample in group j. N and \(Pr(T_j)\) however are population parameters denoting the total population of possible adopters and \(Pr(T_j)\) the unconditional probability of adoption in period j. These quantities are not observed in the sample (we cannot simply assume that the ratio of adopters to non-adopters from a short-run program corresponds to the respective population adoption ratios).

In order to avoid controversies over population priors we rely on a stronger functional form assumption and assume that \(Pr(T_{i}=1|x_i)\) can be written in a multiplicative intercept form (Hsieh et al. 1985). The logit model is a particular example of the multiplicative intercept form and thus we assume that:

$$\begin{aligned} Pr(T_{i}=1|x_i)=\frac{exp(c_1+x'_i\beta _1)}{1+exp(c_1+x'_i\beta _1)}, \end{aligned}$$

where \(\beta _1\) is a parameter vector which measures the extent to which the observed covariates explain adoption. Note that the coefficient \(\beta _0\) is normalized to 0. Thus, the results can be easily interpreted. For each covariate of interest a positive coefficient in \(\beta _1\) tells us that a particular regressor makes it more likely that a household with that characteristic will adopt. The estimated intercept coefficient, \(c_1\), is however inconsistently estimated and is a function of the unknown parameter, \(\mu \). This imposes some restrictions as it prevents us from computing marginal effects without imposing out-of-sample priors on this unobserved parameter.

With these econometric subtleties in mind, let us now turn our attention to Table 7. This table presents estimates of the slope coefficients of the adoption equation for different specifications with an increasing number of explanatory demographics. The results are very similar across different specifications. All else equal, adopters in our sample are younger, wealthier, and live in homes with a smaller number of inhabitants. The first two attributes act as one might expect. Since the adopters enrolled through the PG&E website, this shows that younger households are more likely to invest the effort in searching and finding environmental programs. Similarly, wealthier households appear more willing to incur the costs of search and the (albeit small) increase in electricity prices from enrollment. The mechanism for the importance of household size is open to interpretation. Perhaps there are costs to coordinating with many inhabitants. Autonomy may also play a role; one may view it less advantageous to enroll in an energy-related program when the energy use decisions are made by many different people.

Table 7 Logit model of ClimateSmart adoption

Softer household characteristic also appear to be important, consistent with the findings of Costa and Kahn (2013). Adoption is strongly predicted by the Environmental interest variable. This is consistent with the stylized model in Sect. 2 as this variable proxies for \(\delta \), the extent to which a household is aware of the social cost of carbon emissions. It is interesting to note however that Green Living is not a significant predictor for adoption and in fact it has a negative sign. This may indicate complementarities across domains, whereby if a household already is involved to a substantial degree in other environmental activities they are less likely to adopt a new one. An alternative interpretation may have these households preferring to conserve as part of their lifestyle, rather than purchasing conservation in the form of offsets from an external source. As expected, lifestyle variables related to a perceived interest in the outdoors or activities related to wildlife or camping are also important predictors of adoption. Their presence is likely to also be correlated with the degree to which a household perceives carbon emissions and global warming to be a potential utility cost.

The extent to which a household is involved in the community and local charities also appears to be a strong predictor of adoption. This is also consistent with our model as it reflects overall awareness and concern for the local community. By contrast the propensity to contribute substantial amounts to charitable causes is a negative predictor for adoption. This indicates that adoption into the program is not seen as a charitable contribution or expression of altruism. This is not indicative of a contradiction since it is common to think of contributing to the community financially as being very different than contributing time or effort. The propensity for adoption is substantially higher among high income households. It is interesting to note that adoption does not appear to be driven by the level of education. The presence of children is insignificant as a driver of adoption. While the relatively weak impact of children on adoption may also be explained by the higher age in our sample as a consequence of the time series balance requirement, most of these households will have adult children. Perhaps it shows that they discount the welfare of future generations to a substantial degree, although there are a variety of possible explanations. We find no statistically significant differences between renters and owners. This may also be a consequence of our balance requirement since renters are more likely to be excluded from our sample due to their transitory dwelling patterns.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Harding, M., Rapson, D. Does Absolution Promote Sin? A Conservationist’s Dilemma. Environ Resource Econ 73, 923–955 (2019).

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:


  • Carbon offsets
  • Behavioral rebound
  • Green marketing
  • Energy consumption