The advent of the smart grid is sparking interest in intelligent rate design. But while state and federal goals encourage more efficient rate structures, regulatory and political considerations...
Low-Income Reality Check
Evaluating the impact of dynamic pricing.
participants, CL&P conducted enrollment surveys. For those who responded, the enrollment survey divided participants into one of four income categories, with “less than $50,000” being the lowest. The Wood/Faruqui article treats the “less than $50,000” classification as “low income.” However, the “less than $50,000” category might contain households that aren’t facing financial hardship, e.g., a single professional. These households can’t be expected to respond to dynamic pricing in the same manner as, for example, a family of five.
The impact of missing data on interpreting income-related responsiveness in the CL&P pilot is elsewhere acknowledged by one of the Wood/Faruqui article’s authors. In a document filed with the Connecticut Department of Public Utility Control on Feb. 19, 2010, Dr. Faruqui described a regression evaluation of the socio-demographic factors associated with the CL&P Plan-IT Wise pilot:
Furthermore, the Feb. 19, 2010 report concluded that the difference in price response of customers with incomes above and below $50,000 was “not clear.”
CL&P also separately identified a portion of the pilot population as low-income/hardship customers. With regard to these hardship customers, the Wood/Faruqui article states: “[R]esults indicated that hardship customers responded slightly less than the average treatment customer to the PTP rate, although they did still respond. The incremental effect of the PTR rate was similar for hardship and non-hardship customers.”
However, the Feb. 19, 2010 report states that hardship customers were “non-responsive” for time-of-use (TOU) rates. The evaluation of the CL&P hardship cases led the Feb. 19, 2010 report to conclude that “hardship reduces responsiveness.”
When considering the Baltimore Gas & Electric Smart Energy Pricing (SEP) pilot, the Wood/Faruqui article claims that the elasticity of substitution of low-income customers isn’t statistically different from that of higher income customers. However, because the data set associated with the SEP pilot couldn’t determine the income levels of about 27 percent of pilot participants, the Wood/Faruqui article bases its conclusion on a biased subset of data from a customer survey. 4 Faruqui acknowledged the data’s incomplete nature when he prepared an assessment of the BGE SEP Pilot for Baltimore Gas & Electric:
This again highlights the difficulty of projecting the impact of dynamic pricing on low-income customers—consumers are hesitant to report income levels in surveys. This hesitancy leads to missing data and undermines statistical validity of results. This problem points to the importance of studying the behavior of verified low-income consumers.
The California Statewide Pricing Pilot (CSPP) was conducted in 2003. With regard to the income characteristics of customers in the CSPP, the Wood/Faruqui article states: “Overall, high-income households were somewhat more price-responsive than low-income households; however, the difference wasn’t substantial.”
However, a review of the CSPP points to some problems with the underlying data and to verified low-income customers who are less responsive.
The CSPP had three “tracks.” The CSPP’s Track A included an analysis based on customer income. However, according to the Charles River Associates report filed with the California Commission on the CSPP, the Track A component of the CSPP took a selective approach to developing its sample population and reflected the following