Our results point to challenges to achieve diverse representation. Even supportive rhetoric about how important it is can have negative effects unless elites communicate the group implications of policy-making.

Representation rhetoric's effect depends on the issue's racialization. When racialized, it increases support for political representation of racial minorities. When not, representation rhetoric decreases evaluations of the selection process and candidate.

We find that overall commitment to selecting someone from a particular marginalized group has negative effects on the committee's legitimacy and selection process. These negative effects occur regardless of the racialization of the policy issue.

We vary politicians using this rhetoric. We also vary whether the issue the leader selected is tasked with increasing educational outcomes for all Americans or for Black and Latino Americans to close gaps in racial educational outcome.

2) Representation rhetoric. Politicians have also highlighted how important it is to make political institutions represent Americans descriptively.

We look at how people respond to common ways elites discuss diversifying politics. 1) Commitment rhetoric. Politicians of both parties have committed to selecting someone of a particular marginalized background to fill a powerful role (e.g. KBJ, ACB)

Overall, partisan coalitions respond to obstruction differently and that response depends on the gender of the politician due to sorting. We highlight the need to integrate identities into our studies of elite behavior, with respect to both voters and politicians.

While existing lit suggests women to be consensus-builders, we indicate a more conditional effect where some women are more obstructive to reap electoral gains among Republicans. This can explain why we see Republicans, esp Republican women, credit-claiming for obstruction.

We confirm that our experimental dynamics are showing up in real world politicians' behavior. Republicans, especially Republican women, are most likely to offer obstructive motions in the House and discuss obstruction favorably in e-newsletters.

On the other hand, Democrats tend not to reward obstruction or evaluate women and men who obstruct differently. Because of this, Republicans, particularly Republican women, have incentives to obstruct to garner support with Republican constituents.

Using a conjoint experiment, we find that Republicans reward women more for obstruction. When women are perceived to be unobstructive, Republicans favor men. However, as they are seen as more obstructive, women face no penalty among Republicans.

We argue that obstruction is one way politicians can signal their masculinity. Masculinity is most electorally beneficial among Republican respondents. This is particularly true for for women politicians who are assumed to lack masculinity compared to men.

If people prefer compromise, it's unclear why we see rising gridlock and obstruction. It is even more puzzling we see Republicans, particularly Republican women, credit-claiming for obstruction.