Use of algorithm to target fuel poverty help sparks fears

A government decision to use an algorithm to identify recipients of the Warm Homes Discount (WHD) has sparked concerns that it could result in a 2020 school results-style fiasco with eligible households missing out on payments.

The government has said it will use an algorithm to determine households’ eligibility for rebates through the discount scheme based on their estimated energy costs.

Instead of having to apply every year for a rebate through the scheme, which has been increased from £140 to £150, eligible households of working age will be identified by matching data.

Eligible households will be those with high energy costs who receive means-tested benefits or tax credits. These households will be identified by matching a mix of benefits, property characteristics and energy supplier customer data.

Pensioners who are eligible on grounds of low income will continue to receive the discount automatically.

The government has estimated that the reform of the scheme will identify a further 560,000 eligible households, increasing to 47% the proportion of fuel poor households who receive the discount. The government’s move follows concerns that many eligible working age households are not applying for the discount.

However, during a House of Lords debate earlier this week to approve the updated WHD regulations, opposition peers raised fears that use of an algorithm will lead to some households losing out on payments they are entitled to.

Baroness McIntosh of Pickering, who is also president of National Energy Action, raised concerns that it will be “very difficult” to contest the decision not to select working age households for the discount.

Referring to the award of what were widely seen as unfair GCSE and A-level results by an algorithm in 2020, Labour peer Lord Lennie expressed concerns about the risk of “an education-type embarrassment for the government”

He said: “Lots of customers who potentially would benefit from this scheme.. may then miss out.”.

Responding on behalf of the government, energy minister Lod Callanan said that it is working with external experts on the algorithm

He said: “The government will apply a model that estimates households’ energy costs using property characteristic data. These characteristics are property type, property age and property floor area. They allow us to estimate energy costs accurately while, as far as possible, attempting to keep the process simple and transparent.

“This enables us to ensure that low-income households with the highest energy costs receive a rebate and that the vast majority of households receive those rebates automatically.”

The regulations also slash the eligibility thresholds for suppliers participating in the scheme from 50,000 to 1,000 customers, meaning that 99.4% of the market will be covered, said Lord Callanan.