Government ‘analysis’ on equality impacts flawed, but that is no excuse not to act

Like other public bodies, the UK government is required under the Public Sector Equality Duty to assess the impact of specific policies on ‘protected groups’ including by gender, age, race and disability. For the wide-ranging series of welfare reforms since 2010, this has been done largely through equality impact assessments, introduced in the 2010 Equality Act. As one of the authors of this week’s NIESR report  on the equality impacts of recent welfare reforms, in this post I will reflect on the role and quality of the government’s own impact assessments.

In 2012 the then Prime Minister, David Cameron, said that equality impact assessments were “bureaucratic nonsense” and that policymakers should “use their judgement to do the right thing to meet the equalities duty” rather than “tick boxes”. To a certain extent, he was right. Similarly to the now ridiculed sectoral Brexit analyses most, if not all, equality impact assessments seem only to pay lip service to a requirement, rather than genuinely and thoroughly seeking to understand the wide-ranging impacts of welfare reforms.

What are the problems?

Firstly, the equality assessments are very simplistic and contain limited detailed evidence. They usually use DWP claimant data to provide a breakdown of protected groups, showing the proportion of existing claimants who would likely be affected by the reform. In some cases, the average monetary impact is calculated, but very rarely does the analysis go beyond financial considerations such as claimants’ likely responses to the reform, the impact on employment trajectories, living standards, homelessness, mental health, economic dependency within households etc. In addition, the Government has never published any document exploring the cumulative impact on individuals and households of the changes to a range of benefits, nor responded properly to the cumulative impact assessment co-authored by NIESR in 2014.

Secondly, the equality impact assessments are also surprisingly political and build arguments on controversial (non-evidenced) assumptions, particularly about behaviour of certain groups and their position in society structures. For instance, the equality impact assessment for Universal Credit acknowledges the possibility of reduced work incentives for second earners in couples (primarily women), but justifies the priority to improve the work incentives for the main earner by referring to evidence on the negative impact on children in workless households and then comments that “…second earners may be able to reduce or rebalance their house or to leave work. In these cases, the improved ability of the main earner to support his or her family will increase options available for families to strike their preferred work/life balance.” The Women’s Budget Group has raised concerns that this is reminiscent of a return to a ‘male breadwinner model’.

Similarly, the equality impact assessment for time-limiting contributory ESA acknowledges that a higher proportion of women will not be eligible for income-related ESA to compensate for the loss of income from the contributory ESA, but notes that they will be compensated otherwise since “they will generally either have a working partner or capital over £16,000 so they will not be left without income.” This is not backed up by any evidence, including on the proportion of single parents in this group, nor their average savings.

Another problem is the strange choice of comparators. For instance, when assessing housing welfare reforms, impact assessments tend to compare the proportion of affected claimants with the overall Housing Benefit caseload rather than the UK population. This leads many equality impact assessments to conclude that policies are gender neutral, simply because the existing policy framework was already disadvantaging women. See for instance the equality assessment for the bedroom tax, which concludes that “there is no differential impact by gender” or the equality assessment for Universal Credit which concludes that the policy is gender neutral, and that men and women are treated equally under Universal Credit when they “are in the same circumstances.” However, women are often not in the same circumstances. Surely, the fact that women are in a more disadvantaged position prior to the reform and therefore are affected more negatively should be captured in any conclusion about the gendered impact of a policy.

The best we got, and lack of quality is no excuse not to act

Despite their faults, the equality impact assessments remain the single most detailed piece of evidence on the impact of welfare reforms since 2010. Even in their current form, they are invaluable to understand the impact on protected groups, particularly in benefit areas which have received less attention among researchers and media.

Ultimately, the main issue with the equality impact assessments is a political one. The existing evidence on the disproportionate impact on protected groups, whether published by government or research organisation, has never been properly considered or acted upon by government. Our research published this week by the Equality and Human Rights Commission, alongside an updated cumulative impact assessment, clearly shows the disproportionate impact on protected groups, in particular on disabled people and lone parents. These issues need to be urgently addressed, for the government’s commitment to the Public Sector Equality Duty to be taken seriously.