When does identity matter for policy change? A theory of descriptive and substantive representation

Why are some new policy demands represented, while others are not — even if both have widespread support? For example, the US Democratic party made higher education affordability a major section of its 2016 platform, while work-family policies, like paid parental leave and child care, remain largely relegated to the sidelines. In this paper, I propose a theory of identity and representation which explains why some issues go largely unarticulated by political actors, even when large groups want change. My explanation focuses on the ability of groups to gain political representation, and the alignment of group interests with political party incentives. I suggest that descriptive representation (size of a group within an elected body) ought to matter especially when group interests cut across mainstream left-right dimensions, and highlight several reasons why parties are deterred from advancing these interests without members of the group in office. I consider how various group-based demands fit within the model, before providing preliminary evidence about women’s distinct and cross-party preferences for work-family policies that supports the model’s expectations. I conclude that descriptive representation, and measures that support it like quotas, is especially important for substantive policy change on these cross-cutting issues.