How the individual selection of financial models affects the diffusion of solar energy in residential PV markets
By: Varun Rai and Benjamin Sigrin
Acknowledgements
Varun Rai is a professor at the LBJ School of Public Affairs and the Mechanical Engineering Department, and Benjamin Sigrin is a graduate student in the LBJ School of Public Affairs at The University of Texas at Austin. Rai would like to acknowledge support from the Elspeth Rostow Memorial Fellowship and from the Center for International Energy and Environmental Policy (CIEEP) at the University of Texas at Austin. Benjamin Sigrin would like to acknowledge support from the LBJ School of Public Affairs and the Energy & Earth Resources Program at UT Austin. Errors are ours alone.
This article originally appeared in the Environmental Research Letters journal. All material in IOP’s Open Access journals is, unless stated otherwise, published under a Creative Commons Attribution licence (CC-BY).Open Access articles published in Hybrid journals are made available under the same conditions. We used a rich dataset from the burgeoning residential PV market in Texas to study the nature of the consumer’s decision-making process in the adoption of solar technologies. In particular, focusing on the financial metrics and the information decision-makers use to base their decisions, we study how the leasing and buying models affect individual choices and the adoption of capital-intensive energy technologies. Overall, our findings suggest that the leasing model more effectively addresses consumers’ informational requirements and, contrary to other studies, that buyers and lessees of PV do not necessarily differ significantly along socio-demographic variables. Instead, we find that the leasing model has opened up the residential PV market to a new, and potentially large, consumer segment — those with a tight cash flow.
Introduction
Two questions prompted the work in this paper. First, what can be learned from the diffusion of solar photovoltaics for improving existing solar programs and the design of others in newer markets? As policy support for these technologies is waning, this increases the pressure for incentive programs to become more efficient (U.S. DOE 2012; U.S. DOE 2008). Second, what lessons can the residential PV market shed on the individual decision-making process? The scale of capital investment for solar PV is quite high relative to most other household investments. So presumably, the choice to adopt PV forces individuals to consider the (alternative) options more carefully than most investment decisions (Jager 2006). Unpacking the decision to adopt PV, then, might provide insights into the nature of individual decision-making.
Understanding the nature of the decision-making has important practical implications for the design of mechanisms that encourage reduction of greenhouse gas (GHG) emissions from energy use. With 22.2% consumption of primary energy and 21.4% of the total GHG emissions (EIA 2010), the residential sector is a key target for reducing energy demand and GHG emissions. Diffusion of microgeneration technologies, particularly rooftop PV, represents a key option in meeting demand and emissions reductions in the residential sector (U.S. DOE 2012). As different actors have tried to design programs and incentives to spread the adoption of more efficient and environmentally-friendly consumption and generation devices (Taylor 2008), the nature of the individual’s decision-making process has come to sharper focus (Allcott & Mullainathan 2010; Dietz 2010; Drury et al. 2011; Jager 2006; Keirstead 2007; Bollinger & Gillingham 2012). Therefore, the last few years of experience with residential PV provides an early opportunity to refine our understanding of how individual decision-making affects technology diffusion.
Three lines of theory are relevant to this work:
1. Decision-making at the individual level. While the neoclassical microeconomic theory presumes that individual decision-makers are rational and information-prescient, there is increasing evidence that individual decision-makers depart significantly from the neoclassical model (Camerer et al. 2004; Frederick et al. 2002; Gintis 2000; Todd & Gigerenzer 2003; Wilson & Dowlatabadi 2007).
2. Empirical evidence of the use of high discount rates for future returns from energy-saving technologies (Gately 1980; Hausman 1979; Meier & Whittier 1983; Ruderman et al. 1987). Expectations of rapid technological change, information barriers and other non-monetary costs are some of the factors that give rise to the use of high implicit discount rates (Hassett & Metcalf 1993; Howarth & Sanstad 1995). In general, this phenomenon discourages the adoption of technologies whose benefits are spread over a long time horizon. The use of upfront capital subsidies have been proposed as a way to overcome this adoption barrier (Guidolin & Mortarino 2009; Hart 2010; Jager 2006; Johnson et al. 2011; Timilsina et al. 2011;).
3. Business models for accelerating the deployment of technologies by addressing market barriers (Gallagher & Muehlegger 2011; Margolis & Zuboy 2006; Sidiras & Koukios 2004) facing individual decision-makers,in particular the leasing model. Several researchers suggest that the option to lease a technology effectively addresses the high discount-rate problem (Coughlin & Cory 2009; Drury et al. 2011), as well as some of the information failures associated with new technologies (Faiers & Neame 2006; Shih & Chou 2011).
Data
Our analysis uses a new household-level dataset built through two complementary data streams: (i) a survey of residents who have adopted PV and (ii) program data for the same adopters obtained from utilities that administer PV rebate programs. The survey, among other factors, explores why PV adopters made the financial choices they did (say, buy vs. lease), and their own assessment of the attractiveness of their investment (Rai and McAndrews 2012). The survey was administered electronically in Texas during August-November 2011 and received 365 responses from the 922 PV owners contacted.
All survey respondents reported residing in areas of retail electricity choice in Texas (see Supplementary Information for spatial distribution). The mean size of the PV system installed was 5.85 kW-DC and the average age of respondents was 52 years old. The mean household income was between $85,000 and $149,999 and 84.9% reported that at least one member of the household had achieved a college degree or higher level of education. Each of the prior demographics is significantly different from state-wide averages. That is, the survey population was wealthier, older and better-educated than the average Texas resident. No significant difference was found between lessees and buyers of PV on any demographic variable.
Of the 365 responses, we matched complementary program data for 210 respondents. The program data provides several data points, including (i) installed cost of the system, (ii) price and structure of lease payments if the system was leased, (iii) system capacity (kW, DC and AC), (iv) amount of rebates disbursed, (v) aggregate household electricity consumption from the prior year, (vi) retail electricity provider (REP), electric plan, and marginal cost of electricity consumption just prior to PV installation, and (vii) projected annual electricity generated by the system based on orientation, derating factor and geography.
Methodology
Our strategy is to compare the financial metrics that PV adopters used to evaluate their investment decision (reported metrics) obtained through survey (above) with an “objective” assessment of those same metrics (modeled metrics). To enable the comparison, we built a financial model that calculates the expected lifecycle costs and revenues of PV system ownership for the residential buying and leasing business models (NREL 2009; Kollins et al. 2010). Our model is distinct in two ways. First, our unusually comprehensive dataset allows detailed cost and revenue calculations for each respondent (decision-maker). Second, it includes detailed features of household-level electricity consumption, electricity rates, and PV-based electricity generation, including time-of-day and monthly variations. Next, we provide an overview of our methodology.
3.1 Cash-Flow Model
For each PV adopter we calculate a series of monthly-expected costs (Ck) and revenues (Rk) accrued over the lifetime of the PV system, where k is the number of months since the PV system was installed. Therefore, cash flows (CFk) of the investment are:
CFk = Rk – Ck
Using these cash flows we calculate the net present value (NPV) using a 10% annual discount rate, NPV per DC-kW, payback period for each household’s investment, and estimate each individual’s implicit discount rate.
3.2 System Costs
Costs (Ck) have three monthly components: (a) system payments (Csystemk) — either lease payments or loan payments when financed and a down payment as appropriate, (b) operations and maintenance costs (CO&Mk), and (c) cost of inverter replacement (CInverterk) where:
Ck= Csystemk+CO&Mk+CInverterk
System payments for buyers comprise a down payment in the first period and loan payments if the system was financed. The net system cost is the installed cost less the utility rebate reported in the program data less applicable federal tax credits. We assume that: (i) buyers will make periodic operation and maintenance-related (O&M) expenses equivalent to 0 to 0.75%/year-1 of the system’s installed cost; these O&M costs are expensed equally each month, and (ii) inverters require replacement after 15 years of use and cost $0.70 to 0.95 per DC-Watt. In Section 3.4 we present a set of scenarios that systematically vary these parameters.
Lessees are not obligated to pay O&M or inverter replacement costs as this is a value-added service provided by the lessor (Mont 2004). Therefore, the only costs of ownership incurred are lease payments (up-front payment and monthly lease payments). Within the sample, 69% of lessees paid for their lease entirely through a ‘pre-paid’ down payment, 26% through only monthly payments and 4% through a combination of monthly payments and a down payment. For all leased systems analyzed, we use the actual lease payments being made by the lessees.
3.3 System Revenue
PV systems generate value by reducing owners’ electricity-bill expenses during the life of the system. Therefore, the difference between electric bills the owner would have incurred without the system (BAU bill) and those with the PV system (PV bill) is effectively a monthly stream of revenues (Rk). The value of these revenues depends on the structure and rates of both bills. Our model forecasts these revenues over the system’s lifetime.
3.3.1Electricity Consumption and Generation Profiles
Two central factors in the PV value proposition are seasonal and hourly variations in the system’s generation and the household’s consumption of electricity. For both factors, we use each respondent’s historic annual consumption and expected annual system production (kWh) as reported in the program data, but not individual consumption or generation patterns. To simulate these hourly and seasonal variations we used load profiles published by the Electricity Reliability Council of Texas (ERCOT) of average residential consumption patterns in north-central Texas in 2010 (ERCOT 2010) and a PV generation profile for the Dallas-Ft. Worth area taken from the PVWATTS model created by the U.S. National Renewable Energy Laboratory (NREL 2011).
Furthermore, we assume that patterns and quantities of electricity consumption are invariant over the lifetime of the PV system. This is not a robust assumption per se, since we do not capture household-level patterns of consumption that differ from the ERCOT profile or that evolve over time. But, since the goal is to compare the objective and reported financial metrics, this assumption is robust enough for our analysis because any variations in consumption profiles will largely cancel out in the revenue calculations.
3.3.2 Electricity Rates
Within the ERCOT deregulated electricity market customers freely choose retail electricity service among providers with varying rates and bill structures (TECEP 2012). An important factor is whether their Retail Electricity Provider (REP) offers a plan that credits any moment-to-moment excesses of PV generation over consumption outflowed to the grid (Darghouth et al. 2011; Mills et al. 2008). Unlike many retail-choice states, the ERCOT market does not mandate that REPs provide credits for these ‘outflows’ (PUCT 2012). Current practice is for REPs to credit outflows at a rate below the marginal price of electricity.
While it is tempting to assume that consumers will select electricity plans which offer the highest value for their PV system, it is not obvious what depth of information finding and analysis decision-makers go through to determine which REP provides this greatest value (Conlisk 1996; Fuchs & Arentsen 2002; Gigerenzer & Todd 1999; Goett et al. 2000; Roe et al. 2001; Tversky & Kahneman 1974). We account for this dilemma through a set of scenarios, discussed next.
4.1 Installed Cost and Cost of Ownership
Installed costs ($W-1) of leased systems (Mean = 8.3, Std. dev. = 0.53) were significantly more than those of bough systems (Mean = 6.2, Std. dev. = 1.4), and the mean differences were highly significant (t(201) = 16.08, p < 0.001). This corroborates similar installed cost differences for bought and leased systems nationally (Barbose et al. 2012). As discussed in Section 3.2, recall that while buyers’ cost of ownership is the installed cost less applicable rebates, the installed cost is generally not reflective of the lessees’ cost of ownership, which are only their lease payments. Surprisingly, the mean lessees’ costs of ownership ($0.70 W-1) were substantially less than those of buyers ($2.64 W-1)1. Accordingly, we found that lessees had a statistically significant greater NPV per capacity ratio (NPV/DC-kW) than buyers in all but Scenario 5 (Figure 1; only baseline scenario shown).
How is it possible that leased systems are installed at higher costs than purchased systems, but that lessees face a lower cost of ownership than the equivalent bought system? As others have noted (for example see, Barbose et al. 2012), the installed cost reported to state and utility PV incentive programs is often the ‘fair market value’, or the appraised value, reported when applying for the 1603 Treasury Cash Grant or Federal ITC. Since the benefits of both the 1603 Treasury Cash Grant and tax benefits from MACRS increase with the appraised value of the system, it is plausible that some leasing companies might be inflating the appraised value — at least the incentive to do so clearly exists. Indeed the SEC and IRS recently began an investigation of several leading leasing firms to determine if the true fair market value of installed PV systems were materially lower than what the firms had historically claimed (SEC 2012). If proven true, one implication of this financial strategy would be that since additional system costs and company profits are recouped through the tax structure, leasing companies adopting such strategies would be able to offer lower rates to their customers (the lessees). The fact that we indeed find the cost of leasing PV systems (by the lessees) to be much lower than the cost of buying PV systems lends some support to the hypothesis that some leasing companies might be employing such financial strategies.
Therefore, we tentatively explain lower lessees’ costs of ownership through the following mechanisms: (i) maximization of federal tax benefits by leasing companies (lessors) through the financial strategy described above; (ii) in the current policy environment, lessors are able to access additional financial incentives that buyers cannot access, particularly accelerated depreciation (Bolinger 2009; Coughlin & Cory 2009); (iii) economies of scale present in the operation of a larger fleet of leased systems; (iv) ability for lessors to raise capital at a lower cost, which would increase their leveraged return on capital; and (v) since the lease contracts are typically only 15 to 20 years as compared to the generally reported lifetime of PV panels of 20 to 25 years, leased systems will likely have some residual value. In theory, the lessors could recoup the residual value at a later date, which would allow them to offer the leased systems at lower rates today. All of these mechanisms would lower costs faced by lessors and thereby reduce the size of the lease payments required to achieve a given rate of return. In a competitive leasing market, these mechanisms would translate into lower costs faced by lessees, just as we find. A deeper explanation of these aspects would require financial analysis of the leasing companies’ balance sheets, which is beyond the scope of this paper.
If leasing is financially more attractive, why don’t more adopters choose to lease? For many, the option didn’t exist. 73% of buyers reported not having the option to lease when making their decision. There is also evidence in the literature of conspicuous consumption for novel ‘green’ technologies (Dastrop et al. 2011; Sexton 2011); under this paradigm, consumers could derive additional utility from the status gained by owning, rather than leasing, their systems. Residence uncertainty was not a factor, as each group reported a similar (10-15 years) period that they expected to continue living in their homes. Finally, a majority of PV adopters who had the option to either buy or lease a PV system, but chose to buy, report concerns about potential difficulties with the leasing contract as a factor in their decision to buy. Considering all these factors, we conclude that buyers who did have the option to lease, but chose to buy, had adequate cash-flow such that they preferred the contractually simple buying option, even though the leasing option is nominally cheaper.

Figure 2: Comparison of reported and modeled payback period in scenario 3. Mean difference between modeled and consumer payback period: Buyers = 7.1 yrs-1; Leasers = 1.1 yrs-1.
4.2 Payback Period Comparison
Consistent with previous research (Camerer et al. 2004; Kempton & Montgomery 1982; Kirchler et al. 2008), the majority of respondents (66%) reported using payback period to evaluate the financial attractiveness of their investment as opposed to NPV (7%), internal rate of return (27%), net monthly savings (25%), or other metrics (6%). 10% made no estimate of the financial attractiveness. Respondents also reported the values of the metrics they used. These responses allow us to compare reported metric values (reported) to the values individually generated from the financial model (modeled) (Figure 2; only baseline scenario shown).
For buyers, Scenario 4 minimized the average absolute difference between reported and modeled payback period (M = 2.6 years, SD = 2.4), followed by Scenario 5 (M = 3.1, SD = 1.9). For lessees, Scenario 3 (M = 1.1, SD = 0.7) was the best fit, followed by Scenario 2 (M = 1.296, SD = 0.704). Scenario 1 was a poor fit overall. This suggests that buyers assumed parameters similar to those of Scenario 4 when evaluating their investment. That is, buyers were optimistic when assessing the likely revenues and costs associated with their investment decision. By the same argument, lessees were more realistic and precise when making their investment decision. This is consistent with the fact that lessees receive much of this financial information from leasing companies, who use detailed and sophisticated financial models.

Figure 3: Distribution of implied NPV/kW for buyers and leasers; Difference of mean is not significantly different than zero.
4.3.1 Implied Discount Rate
For all calculations of NPV reported above a 10% annual discount rate was assumed. In this section we present discount rates calculated separately for each individual respondent. Specifically, we first determine each respondent’s implied NPV and then back-calculate their discount rate using the implied NPV and their modeled cash flows. To determine the implied NPV, respondents were asked on a 5-point Likert-scale how strongly they agreed with the following five statements: (i) “I would not have installed the PV system if it had cost me $1,000 more”… (v) “I would not have installed the PV system if it had cost me $5,000 more.” One expects respondents to increasingly agree that they would not have installed the PV system as the price increased. The above question estimates the respondent’s implied NPV by extrapolating how much more the respondent would have paid before becoming indifferent to purchasing the system or forgoing the investment (Figure 3).
Of the 210 respondents in our dataset, 92 responses were excluded from these calculations — 69 whose implied NPV was outside the range tested ($0 to $5,000), 7 responses which implied an increasing willingness to pay, and 16 non-responses. Of the excluded respondents, 55 respondents indicated they would have been willing to pay at least $5,000 more for their system, of which 76% were buyers and 24% leasers. That is, a significant percent of the sample (26.2%) did assign a positive value to their investment, yet were not captured within this calculation because of insufficient data. In the end, there are 81 buyers and 37 lessees remaining for the discount-rate analysis reported in this section.
Using the implied NPV, we solve for the monthly discount rate (rm) required to equate the respondent’s implied NPV with the cash flows modeled earlier:
The monthly discount rate is then annualized using:
r = 1 + rm – 1
Thus, r represents each respondent’s discount rate implied by their willingness-to-pay and their modeled cash flows. As the cash flows vary with each scenario, implied discount rates also vary with scenarios.
Using baseline (Scenario 3) parameters, the mean discount rate for buyers was 7 ± 5% and for lessees was 21 ± 14% (± 1 σ ) (tables 2 and 3). The calculated implied discount rates are higher in the optimistic scenarios since cash flows increase as the scenarios become more optimistic. Across all scenarios and income levels lessees’ implied discount rates are significantly higher than buyers by 8 – 21%.
It’s important to note a similarity in the timing of leased and bought payments — the majority (69%) of lessee respondents chose to structure their leases as a single ‘prepaid’ down payment, which is similar to the financial structure of a bought system, but significantly smaller in the scale of investment. After taking all incentives into account, for lessees the upfront payment is on the order of $4000 and for buyers it is $15,000 for a 6 kW-DC system. Yet each group expects to receive a similar (normalized) NPV for their investment. That is possible only when these groups have differing cash urgencies. Indeed, in open-ended survey questions, 66.2% of lessees agreed or strongly agreed that tight cash-availability was one of the key factors in their decision to lease, whereas buyers generally did not have this problem. Given that there are little, if any, demographic differences between buyers and lessees, we infer that at this stage in the residential PV market buyers and lessees represent different consumer segments within a similar socio-demographic makeup. Put differently, compared to the average buyer the average lessee is not lower income per se — the majority of the lessees have some cash availability, but not enough to outright buy their PV system.
In general, our point is that within populations with similar demographics, it’s possible that there are variations in disposable income, and those variations are a key factor in ownership model choices. Consistent with a large body of work in the diffusion of innovations tradition (Rogers 2003), our results suggest that there is a hierarchy within the population regarding the adoption of technologies. In early stages of technology diffusion, as is the case with PV now, information (awareness of products, interest in energy, etc.) is the precursor, which is more likely to be found in higher income, more educated segments of the population. Within those segments, those with tighter cash flows opt for leasing if the option is available. Thus, the leasing model appears to be especially effective in the early stages of a technology’s diffusion, as it unlocks the cash-strapped but information-aware segments of the market. Put differently, the leasing model accelerates the early adoption stage of a technology’s diffusion, thereby quickly establishing a wider base upon which later adoption can build.
4.4.1 Discount Rate and Income
Previous literature starting with Hausman (1979) suggests that an inverse relationship exists between household income and consumer discount rate. That is, poorer consumers have more urgent needs for their cash than wealthy ones. At higher incomes, where one has a greater degree of spare income, the rate of return of investments (and hence, their discount rate) should converge to market returns. Our results are mixed in regard to these earlier findings.
A one-tailed t-test comparing the difference in mean discount rate among income groups for the baseline scenario was performed using the hypotheses Ho: DR1 = DR2, Ha: DR1 ≥ DR2, and Ho: DR2 = DR3, Ha: DR2 ≥ DR3, where DR1 is the mean implied discount rate for income group 1 and so on. This test was performed for both income pairs (DR1 ≥ DR2 ≥ DR2 ≥ DR3) since we expect the implied discount rate to monotonically decrease with income.
Even with a 90% confidence interval, we did not find a statistically significant relationship between income and discount rate for either buyers or lessees. We explain this discrepancy with two reasons. First, small sample size, particularly in the leasing sample, reduced our test’s statistical power. Second, both groups exhibit characteristics typical of early adopters — wealthier, more educated,
etc. These characteristics could negate the relationship between income and discount rate for products in settled markets as early adopters typically derive additional utility from adopting new technologies beyond financial benefits (Faiers et al. 2007; Labay & Kinnear 1981; Rogers 2003). In agreement with previous literature, we do find that discount rates for buyers in the conservative, baseline, and optimistic scenarios (Scenarios 2-4) ranges between 7 and 13%, which is close to market returns. This also supports our finding that buyers of PV systems are in a relatively comfortable cash-flow position.
Conclusion
We have studied the economics of the decision-process of individual consumers, particularly their decision to buy or lease a residential PV system. Consistent with several other studies, we find that a majority of PV adopters used payback period — not net present value (NPV) — as the decision-making criterion. We also find that owing to the peculiarities of financing and incentive mechanisms, the pre-rebate installed costs of leased PV systems are significantly higher than the bought systems, yet lessees end up paying nominally much lower amounts than buyers of PV. We calculate individual-level discount rates across a range of scenarios, finding that buyers employ discount rates 8 to 21% lower than lessees. Those who lease typically have a tighter cash flow situation, which, in addition to less uncertainty about technological performance, are the main reasons for them to lease. As we do not find any significant variation between buyers and lessees on any socio-demographic dimension (income, age, etc.) this suggests that the leasing model is making PV adoption possible for a new consumer segment — those with a tight cash-flow situation. As the diffusion of PV spreads to lower-income households, who generally experience tighter cash-flow than wealthier households, this implies that, ceteris paribus, moving forward the leasing model will likely be the predominant form of PV adoption. From this perspective, the leasing model has opened a new market segment at existing prices and supply chain conditions — and represents a business model innovation. SPW
1 Note that the upfront cost-of-ownership does not reflect the operational life of PV systems or their performance over that lifetime. In general, most analyses assume an operational life for PV systems of 20-25 years, which is applicable to buyers of PV systems. Lease contracts typically terminate after 15-20 years. So the difference in the upfront cost-of-ownership of bought vs. leased systems should be put in this context. However, as discussed below, NPV calculations incorporate this difference in the length of cash flows.
References
Allcott, H. and Mullainathan, S, (2010). Behavior and energy policy. Science, 327, pp 1204 -1205.
Barbose, G, Darghouth, N., and Wiser, R. (2012). Tracking the sun V: A historical summary of the installed price of photovoltaics in the United States from 1998 to 2011. Environmental Energy Technologies Division, Lawrence Berkeley National Laboratory. LBNL-5919E. November 2012.
Bergmann, A., Hanley, N., and Wright, R., (2006). Valuing the attributes of renewable energy investments. Energy Policy, 4 (9), pp 1004-1014.
Bolinger, M.A. (2009). Full steam ahead for PV in US homes?, Renewable Energy Focus, 9(7), January–February 2009, pp 58-60, ISSN 1755-0084, 10.1016/S1755-0084(09)70041-9.
Bollinger B and Gillingham, K 2012 Peer effects in the diffusion of solar photovoltaic panels. Marketing Science vol 31 no 6 900 – 912 doi:10.1287/mksc.1120.0727
Camerer, C., Loewenstein, G., and Rabin M, eds, (2004). Advances in Behavioral Economics. Princeton, NJ: Princeton University Press.
Conlisk, J., (1996). Why bounded rationality? Journal of Economic Literature, 34, pp 669–700.
Coughlin, J. and Cory, K., (2009). Solar Photovoltaic Financing; Residential Sector Deployment. National Renewable Energy Laboratory Technical Report. NREL/TP-6A2-44853, March 2009.
Darghouth, N.R., Barbose, G., and Wiser, R, (2011). The impact of rate design and net metering on the bill savings from distributed PV for residential customers in California, Energy Policy, 39(9), September 2011, pp5243-5253, ISSN 0301-4215, 10.1016/j.enpol.2011.05.040.
Dastrop, S., Zivin, J.G., Costa, D.L., and Kahn, M.E., (2011). Understanding the solar home price premium: electricity generation & green social status. NBER Working Papers 17200, National Bureau of Economic Research, July 2011.
Dietz, T., (2010). Narrowing the US energy efficiency gap. Proceeding of the National Academy of Sciences, 107(37), September 2010. pp 16007-16008
Drury, E. et al., (2011). The transformation of southern California’s residential photovoltaic market through third-party ownership, Energy Policy, 42(3), pp 681-690.
Energy Information Agency, U.S. (EIA), (2010). Annual Energy Review. Accessed 8 July 2011 at: www.eia.gov/totalenergy/data/annual/index.cfm
Electricity Reliability Council of Texas (ERCOT). (2010). ERCOT Backcasted (Actual) Load Profiles – Historical. Retrieved September 15th, 2011, from http://www.ercot.com/mktinfo/loadprofile/alp/
Faiers, A., Neame, C., and Cook, C., (2007). The adoption of domestic solar-power systems: do consumers assess product attributes in a stepwise process? Energy Policy, 35(6) , 3418-3423.
Faiers, A. and Neame, C. (2006). Consumer attitudes toward domestic solar power systems. Energy Policy, 34(14), September 2006.
Frederick, S., Loewenstein G. and O’Donoghue ,T., (2002). Time discounting and time preference: a critical review, Journal of Economic Literature, American Economic Association, 40(2), pp 351-401.
Fuchs, D.A. and Arentsen, M.J. (2002). Green electricity in the market place: the policy challenge. Energy Policy 30(6). pp. 525–538.
Gallagher, K.S. and Muehlegger, E. (2011). Giving green to get green? incentives and consumer adoption of hybrid vehicle technology. J. of Environmental Economics and Management, 61, pp. 1-15.
Gately, D, (1980), Individual discount rates and the purchase and utilization of energy-using durables: comment, Bell Journal of Economics, 11(1), p. 373-374.
Gigerenzer, G. and Todd, P., (1999). Simple heuristics that make us smart. Oxford: Oxford University Press.
Gintis, H. (2000). Beyond homo economicus: evidence from experimental economics. Ecol. Econ. 35, 311–22.
Goett, A.A., Hudson, K., and Train K.E. (2000). Customers’ choice among retail energy suppliers: the willingness-to-pay for service attributes, The Energy Journal, International Association for Energy Economics, 0(4), pages 1-28.
Guidolin, M., and Mortarino, C., (2009). Cross-country diffusion of photovoltaic systems: modeling choices and forecasts for national adoption patterns. Technological Forecasting and Societal Change, 77(2), July 2009, pp 279-296. http://dx.doi.org/10.1016/j.techfore.2009.07.003
Hart, D.M. (2010). Making, breaking, and (partially) remaking markets: state regulation and photovoltaic electricity in New Jersey. Energy Policy, 38, pp. 6662-6673.
Hausman, J. (1979). Individual discount rates and the purchase and utilization of energy-using durables, Bell J. Econ. 10(1), pp. 33– 54.
Hassett, K. A. and Metcalft, G.E. (1993). Energy conservation investment: do consumers discount the future correctly?, Energy Policy, 21(6), June 1993, pp 710-716.
Howarth, R.B. and Sanstad, A.H. (1995). Discount rates and energy efficiency. Contemporary Economic Policy, 13(3). July 1995, pp 101.
Jager, W. (2006) Stimulating the diffusion of photovoltaic systems: a behavioural perspective, Energy Policy, 34(14), September 2006, pp 1935-1943, ISSN 0301-4215, 10.1016/j.enpol.2004.12.022.
Johnson, K. et al (2012). Lessons learned from the field: key strategies for implementing successful on-the-bill financing programs. Energy Efficiency 5(1) February 2012. p. 109 – 119 DOI: 10.1007/s12053-011-9109-7
Keirstead, J. (2007) Behavioural responses to photovoltaic systems in the UK domestic sector, Energy Policy, 35(8), August 2007, pp 4128-4141, ISSN 0301-4215, 10.1016/j.enpol.2007.02.019.
Kempton, W. and Montgomery, L. (1982) Folk quantification of energy, Energy, 7(10), October 1982, pp 817-827, ISSN 0360-5442, 10.1016/0360-5442(82)90030-5.
Kirchler, E., Hoelzl, E., and Kamleitner B. (2008). Spending and credit use in the private household, Journal of Socio-Economics, 37(2), pp 519-532.
Kollins, K., Speer, B., and Cory, K. (2010). Solar PV project financing: regulatory and legislative challenges for third-party PPA system owners. National Renewable Energy Laboratory. Technical Report 6A2-46723. February 2010.
Labay, D.G. and Kinnear, T.C. (1981). Exploring the consumer decision process in the adoption of solar energy systems. Journal of Consumer Research, University of Chicago Press, 8(3), pp271-78, December.
Laitner, J.A. et al. (2003). Room for improvement: increasing the value of energy modeling for policy analysis, Utilities Policy, 11, 87-94.
Margolis, R. and Zuboy, J. (2006). Nontechnical barriers to solar energy use: review of recent literature. National Renewable Energy Laboratory Technical Report. NREL/TP-520-40116, September 2006.
Meier, A. K. and Whittier, J. (1983), Consumer discount rates implied by purchases of energy-efficient refrigerators, Energy, 8(12), pp 957-962, Dec 1983.
Mills A., Wiser, R., Barbose, G. and Golove, W. (2008). The impact of retail rate structures on the economics of commercial photovoltaic systems in California, Energy Policy, 36(9), September 2008, pp3266-3277.
Mont, O. (2004). Product-service systems: panacea or myth? Ph.D. Dissertation, Lund University, Sweden
National Renewable Energy Laboratory. (2011). About PVWATTS Viewer. NREL: PVWATTS, Retrieved October 3rd, 2012 from http://www.nrel.gov/rredc/pvwatts/about.html
National Renewable Energy Laboratory (2009). Solar leasing for residential photovoltaic systems, NREL/FS-6A2-43572, http://www.nrel.gov/docs/fy09osti/43572.pdf, accessed October 7th, 2012.
Public Utility Commission of Texas (PUCT). (2012), Arrangements between qualifying facilities and electric utilities, §25.242, 2012.
Rai, V. and McAndrews, K., (2012). Decision-making and behavior change in residential adopters of solar PV. Proceedings of the World Renewable Energy Forum, Denver, Colorado, May 2012.
Roe, B., Teisl, M. F., Levy A., and Rissell, M., (2001). US consumers’ willingness to pay for green electricity. Energy Policy, 29(11), pp 917-925.
Rogers, E.M., (2003). Diffusion of innovations, 5th ed. New York: Free Press 2003.
Ruderman, H., Levine, M. and McMahon, J., (1987), The behavior of the market for energy efficiency in residential appliances including heating and cooling equipment, The Energy Journal, 8(1), pp. 101-124.
SEC (2012). Security and Exchange Commission Form S-1, SolarCity Corporation. October 5, 2012. Available from SEC online database. Accessed December 13th, 2012.
Sidiras, D.K. and Koukios, E.G. (2004). Solar systems diffusion in local markets. Energy Policy, 32(18). pp. 2007–2018.
Sexton, S.E. (2011). Conspicuous conservation: the Prius effect and willingness to pay for environmental bona fides Working Paper. http://works.bepress.com/sexton/11. Accessed October 7th, 2012.
Shih, L.H., and T.Y. Chou, (2011). Customer concerns about uncertainty and willingness to pay in leasing solar power systems. International Journal of Environmental Science and Technology, 8(3), pp 523-532.
Todd, P.M. and Gigerenzer, G., (2003). Bounding rationality to the world. J. Econ. Psychol. 24, pp 143–65.
Tversky, A. and Kahneman, D. (1974). Judgment under uncertainty: heuristics and biases. Science 185, pp1124–31.
Taylor, M. (2008). Beyond technology-push and demand-pull: lessons from California’s solar policy, Energy Economics, 30(6), November 2008, pp 2829-2854, ISSN 0140-9883, 10.1016/j.eneco.2008.06.004.
Texas Electric Choice Education Program (TECEP). (2012). Retrieved October 7th, 2012 from http://www.powertochoose.org/
Timilsina, G.R., et al. (2011). A review of solar energy: markets, economics, and policies. Work Bank Policy Research Working Paper No. 5845, October 2011.
U.S. Department of Energy (2012). Sunshot vision study. Washington, D.C.: Government Printing Office. http://www1.eere.energy.gov/solar/pdfs/47927.pdf
U.S. Department of Energy (2008). Multi Year Program Plan 2008 – 2012. Washington, D.C.: Government Printing Office.
Wilson, C. and Dowlatabadi, H. (2007). Models of decision making and residential energy use. Annual Review of Environment and Resources,
32: 169-203, DOI: 10.1146/annurev.energy.32.053006.141137
Thank you for your analysis. Unfortunately it appears that some of the data that you used to draw your conclusion is flawed. For instance: In a real world analysis, there’s absolutely no such thing as a $0 down solar lease or PPA. A requirement of both of these financing programs is that you agree upfront to give the leasing or PPA company your 30% federal tax credit which is worth thousands of dollars as well as any other financial incentives. The CEC currently reports an average system price of $6.07 a watt. Even if you subtract a gracious 50 cent per Watt rebate, that’s $5.57 per Watt. At that price a 6 kW solar system would yield a federal tax credit of $10,026.00.
Second the mean value that you indicated, concerning a price per watt of $6.20, even with the $1.40 per watt deviation for a purchased system is also flawed. Today, a consumer can purchase and own an installed, name brand, grid tie solar system for only $3.15 a watt before incentives, while keeping all of the financial incentives. $6.20 per watt is more reflective of the much higher pricing that the leasing and PPA companies are quoting, not the real world, fair market values that exist in today’s market.
Third, your statement regarding system cost: “System payments for buyers comprise a down payment in the first period and loan payments if the system was financed.” is also flawed. Today, there exists both an FHA $0 down solar loan with tax deductible interest and a $0 down conventional loan that requires no collateral and no fees with interest rates as low as 4.99%.
Fourth, unless I overlooked this, it appears that you failed to mention anything concerning property value increase when a solar system is purchased (an asset) for the potential homebuyer, versus a leased system’s payments (a liability). In fact real estate brokers are currently warning that they are finding it far more difficult to sell a home with a leased solar system than it is with a purchased solar system which can actually raise the selling price.
First: Since the 30% federal tax credit reduces the solar installer’s cost it allows the installer to put the system on at a lower rate per kW. If you believe they just keep the 30% to pad their profits then consider how competitive the solar industry is and TRY to imagine a world where they are not cutting significantly into their 30% federal tax credit in order to beat a competitor… Really, this is a free market isn’t it?
Second: There is a lot of cost associated with the maintenance of the solar system, especially the inverters which, in real world terms, will fail in about 12 years which is after the 10 year warranty an owned inverter usually has (some companies back their purchase solution well beyond the manufacturer’s warranty). The inverter is $3k to $5k and there is labor cost to replace it to boot. Also, a purchase does not typically include insurance (but sometimes does) and production guarantees (but sometimes does). Consult your insurance agent to determine your increased monthly insurance cost. The upshot is this, with a leased system from the biggest installer in the country every headache and expense is taken care of, including the massive cost of inverter replacement ’round about the twelfth year. This REALLY adds up if there is more than one inverter.
Third; 4.99+% interest? Compounded? Really? Do the math before you buy- Not to say there is no way a cash purchase can make more sense but, at 4.99+% interest compounded yearly, can a cash purchase, if financed (which really folks, we just got out of a credit crisis [sort of]: is anything that is financed really a cash purchase?) compete too well with 2.9% fixed rate increase per year? If 2.9% fixed increase on the low cost per kWh sounds high the average rate of increase per kWh from utilities has been about 4.8% since electricity has been sold by the kWh.
Fourth: Berkeley Labs did a study that shows homes with solar- however it is on there, leased, owned, or PPA’d commands an average of $10k-$16K more, period. That is because each method of acquiring solar reduces the cost of electricity, period (unless you sign a contract that has a “raise rates at discretion” clause). It’s the lower cost to use electricity that drives up selling prices. Do check your contract to ensure you can transfer your Solar to the next homeowner without any fees of any kind (I’ve seen fees up to $20,000 in the terms). The largest solar installer only requires a 680 FICO credit score to transfer the contract to the new owner. No lien, no impact on income to debt ratios, and no boogeymen in the bushes. With a $0 down installation you do not have to live in your home an average of 5-12 years before you a) put real money back in your pocket, b) can sell your home at a high enough price to recoup the original investment, and c) have to wrestle with the ethical dilemma of whether or not to disclose to the new buyer that the inverter’s warranty may be about to expire and will cost them a mint to replace it when it dies.
I do work for the largest solar installer and we offer purchase, lease and PPA options. I don’t care how my customer chooses to acquire cleaner, more affordable power but it seems curious to me that so many other solar companies spend so much time and money trying to convince others that options they can’t or don’t offer are bad instead of focusing on what makes them better.
If you leased a system for 20 years and want to sell your house in year 5 – doesn’t that mean that your potential home buyer is inheriting a LIABILITY? if you own your own system – the buyer is getting an ASSET.