Berkeley Lab reviews participant outcomes in Pay-As-You-Save® (PAYS®) programs

April 18, 2024

Pay-As-You-Save (PAYS®) programs have attracted considerable interest as a vehicle to overcome access and affordability barriers to energy efficiency improvements in buildings. These programs enable participating households to receive energy efficiency upgrades without taking on debt by collecting repayment through a tariff attached to a utility meter. PAYS® project approval requires that a project’s energy savings cover the monthly tariffed charge; the programs do not screen participants for credit score or debt-to-income thresholds.

In a new report, Berkeley Lab delineates key design elements of PAYS® programs, describes participant demographic and socio-economic characteristics, and reviews participant energy financial outcomes. The analysis considers PAYS® programs sponsored by five rural electric cooperatives operating in Arkansas, Kansas, Kentucky, North Carolina and Tennessee.

PAYS® programs are active in disadvantaged communities
We find that the average participant lives in an area with household incomes ($51,000-73,000) well below the national average of $89,000. Additionally, participants in four of the five programs in our study live in areas where 16% or less of the population has a bachelor’s degree or higher, much lower than the 32% national average.

PAYS® projects often include customer copays that enable project to pass the PAYS® cash-flow test
Because PAYS® programs require that monthly expected annual energy savings exceed tariffed charges, some projects may not qualify without incentives or upfront customer payments (copays) that reduce the amount recovered through the tariffed charge. In the programs in our study, the share of projects that involve copays ranges from 15-74% by program, and the average copay ranges from $1800-2700 by program. Projects with HVAC and insulation measures are 22% and 8% more likely to have copays and projects with air and/or duct sealing are 12% less likely to have copays. We also find that projects operated by Midwest Energy in Kansas were more likely to have copays than projects in other programs.

Electricity and gas usage fell in most participating Midwest Energy PAYS® homes
We acquired electricity and gas usage data from households participating in Midwest Energy’s How$mart program and estimated weather-normalized changes in monthly electricity and gas usage for these households. Across all projects, we find that electricity and gas usage fell by 15% and 26% on average, respectively, in the year following energy efficiency upgrades financed by the program. At the individual project level, a large majority of Midwest Energy projects reduced electricity (77%) and gas (89%) usage. We are not surprised to find that some households increase electricity or gas usage. While we normalize for weather, we are not able to account for many other factors that influence energy usage, such as the economy, the installation of new appliances or equipment, changes in occupancy or other behavioral changes. These factors may influence usage both before and after a PAYS® project for reasons unrelated to the energy efficiency project installed. 

Figure showing usage


About half of Midwest Energy PAYS® participants have reduced energy bills
We estimate illustrative monthly cash flows for Midwest Energy PAYS® participants using the weather-normalized electricity and gas savings and electricity and gas price scenarios built from price data reported by Midwest Energy. About half of participants appear to reduce their energy costs enough to offset the tariffed charge, saving money in a normal weather year. Energy bill reductions are highest in summer and winter when electricity and gas usage are high for space cooling and heating respectively. This seasonality in bill reductions smooths monthly utility bill expenditures, which is likely valuable to households that manage costs on a monthly basis.

In our view, the energy and cost impacts of the Midwest Energy projects are within reasonable expectations. Given the inherent variability in household energy consumption and cost, volatility in energy prices, and the influence of non-project factors, we would not expect that every participating household’s net energy costs would go down.

For additional results and discussion of the participants outcomes in PAYS programs, see our technical report.