Hostname: page-component-6898cfcb55-cbk24 Total loading time: 0 Render date: 2026-11-11T07:49:20.779Z Has data issue: false hasContentIssue false
  • English
  • Français

How do Public Officials Learn About Policy? A Field Experiment on Policy Diffusion

Published online by Cambridge University Press:  10 February 2021

Miguel M. PereiraOpen the ORCID record for Miguel M. Pereira[Opens in a new window]
Miguel M. Pereira*
Affiliation:
University of Southern California, Los Angeles, CA, USA
*
Corresponding author. E-mail: m.pereira@usc.edu
Rights & Permissions [Opens in a new window]

Abstract

Prior research suggests that partisanship can influence how legislators learn from each other. However, same-party governments are also more likely to share similar issues, ideological preferences and constituency demands. Establishing a causal link between partisanship and policy learning is difficult. In collaboration with a non-profit organization, this study isolates the role of partisanship in a real policy learning context. As part of a campaign promoting a new policy among local representatives in the United States, the study randomized whether the initiative was endorsed by co-partisans, out-partisans or both parties. The results show that representatives are systematically more interested in the same policy when it is endorsed by co-partisans. Bipartisan initiatives also attract less interest than co-partisan policies, and no more interest than out-partisan policies, even in more competitive districts. Together, the results suggest that ideological considerations cannot fully explain partisan-based learning. The study contributes to scholarship on policy diffusion, legislative signaling and interest group access.

Keywords

policy learningpartisanshipfield experimentpolicy diffusionAmerican politics

Information

Type
Letter
Information
British Journal of Political Science , Volume 52 , Issue 3 , July 2022, pp. 1428 - 1435
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution, and reproduction in any medium, provided the original work is properly cited.
Copyright
Copyright © The Author(s), 2021. Published by Cambridge University Press

When facing a new issue in their constituency, elected officials can consider policy solutions that have been successfully implemented elsewhere. This ability to learn from the experience of others can make policy making more efficient, allowing governments to replicate virtuous initiatives while avoiding flawed solutions (Böhmelt et al. Reference Böhmelt2016; Graham, Shipan and Volden Reference Graham, Shipan and Volden2013). However, the prospects of policy learning depend on from whom public officials are willing to learn. If politicians systematically favor (or dismiss) initiatives implemented by a subset of peers, regardless of the specific characteristics of the policy, this process may produce inefficiencies. Bounded policy learning can lead to the proliferation of poor policy solutions. Despite the importance of partisanship in most aspects of political life, few studies have examined its direct role in the diffusion of new ideas. In this project, I explore how representatives use partisan cues to consider which policies to pursue, and how these cues may constrain policy learning.

Prior work established that geographical proximity (Case, Hines and Rosen Reference Case, Hines and Rosen1993), ideological congruence (Grossback, Nicholson-Crotty and Peterson Reference Grossback, Nicholson-Crotty and Peterson2004) and demographic similarities among voters (Simmons and Elkins Reference Simmons and Elkins2004) are associated with learning-based policy diffusion.Footnote 1 Observational studies based on aggregate patterns of policy adoption have played an important role in uncovering diffusion patterns (Graham, Shipan and Volden Reference Graham, Shipan and Volden2013). However, these contributions are unable to isolate what motivates legislators to learn from each other.

To explore how partisanship shapes policy learning, I carried out a collaborative field experiment with a non-profit organization. Welcoming America is an advocacy group that promotes immigrant inclusion policies among American local governments. The experiment was embedded in an outreach campaign conducted by the organization to raise awareness of a new certification program. The email campaign targeted 18,741 local elected officials and included an endorsement by a county government that had recently adopted the policy with support from both parties.Footnote 2 The party affiliation of the officials endorsing the policy was manipulated without deception. Some officials received an endorsement from co-partisans. A second group received an endorsement from members of the other party. Finally, a third group received a bipartisan endorsement.

The study shows that local politicians are systematically more interested in the same certification program when endorsed by co-partisans. The differences are meaningful. Co-partisan endorsements are associated with a 10.7 per cent increase in policy interest compared to out-partisan endorsements. Endorsements by members of both parties attracted no more interest than out-partisan endorsements, and significantly less interest than the co-partisan policy. Nor did bipartisan endorsements attract more interest in politically competitive communities, where voters reward moderate policies (Ansolabehere, Snyder and Stewart Reference Ansolabehere, Snyder and Stewart2001; Carson and Williamson Reference Carson and Williamson2018). Hence, partisan cues do not seem to be used exclusively to filter policies that better reflect the preferences of the constituency. The results are more consistent with a pattern of learning based on in-group favoritism.

Partisanship and Diffusion

A consistent finding in the policy diffusion literature is that legislators are more likely to learn from peers who are more similar to them (Case, Hines and Rosen Reference Case, Hines and Rosen1993). Because partisanship is such an important political characteristic, it is likely to play a role in how officials respond to policies implemented by other governments. Prior observational work finds patterns that are consistent with this relationship (Graham, Shipan and Volden Reference Graham, Shipan and Volden2013). The dissemination of social distancing policies to combat the spread of COVID-19 may be the most recent example of this process (Adolph et al. Reference Adolph2020). However, same-party governments are also more likely to face similar issues, as the constituents who elected them share analogous interests and demands. Hence, co-partisan officials may be inclined to independently try similar policy solutions at the same time. This process of independent experimentation would lead to patterns of policy innovation that are observably equivalent to the results of learning-based diffusion (Volden, Ting and Carpenter Reference Volden, Ting and Carpenter2008). Therefore, observational studies – although compelling – leave out the possibility that the relationship between partisanship and diffusion is epiphenomenal.

Two recent studies have tried to isolate the effect of partisanship on policy learning. Butler et al. (Reference Butler2017) looked at the ideological preferences of local officials. The authors studied the officials’ willingness to learn about a liberal policy and found that the more liberal a politician, the more interested she was in considering the policy. However, this effect was moderated by whether the policy was currently being implemented by co-partisans. The same patterns were uncovered in a separate study for a conservative policy (Butler and Pereira Reference Butler and Pereira2018a). Both studies are based on survey experiments with elected officials. Whether these patterns affect the behavior of representatives once they are in office remains unclear (Barabas and Jerit Reference Barabas and Jerit2010). Individuals derive utility from partisan cheerleading in surveys (Bullock et al. Reference Bullock2013; Butler and Pereira Reference Butler and Pereira2018b). Hence, survey-based evidence can overstate the impact of partisanship. My empirical strategy allows me to overcome these limitations and directly test the behavioral implications of partisan-based learning.

Moreover, to the extent that early adopters' partisanship plays a role, it remains unclear why. Public officials may use partisan endorsements to infer the ideological content of a given policy. As reelection-seeking actors, legislators benefit from pursuing policies aligned with their constituents’ preferences (Canes-Wrone, Brady and Cogan Reference Canes-Wrone, Brady and Cogan2002; Hall Reference Hall2015), even in low-information elections (Hogan Reference Hogan2008). However, most policy issues are complex, and officials have limited resources to process all the relevant information (Zelizer Reference Zelizer2019). Partisan endorsements can be an efficient cognitive shortcut.

In-group favoritism may also explain why representatives rely on partisan cues. Partisanship is a powerful social identity (Barber and Pope Reference Barber and Pope2019). In recent years, researchers have documented a growing divergence in affect toward members of the other party (Iyengar, Sood and Lelkes Reference Iyengar, Sood and Lelkes2012). Partisans are progressively willing to exhibit animus toward and distrust of members of the other party (Abramowitz and Webster Reference Abramowitz and Webster2018). If public officials process new policy information through an in-group/out-group lens, any initiative associated with the other party should be deemed less credible or trustworthy, and generate less interest. This study was not designed to adjudicate between potential mechanisms for partisan-based learning. Still, the results shed some light on the plausibility of these different explanations.

Empirical Strategy

In order to reproduce a process of policy diffusion, I conducted a between-subjects field experiment in collaboration with an advocacy group. Welcoming America is a non-governmental organization that promotes immigrant inclusion policies in local governments. The experiment was embedded in an outreach campaign conducted by the organization in February 2018 that targeted local elected officials. The main goal of the campaign was to disseminate a recently developed certification program designed to provide a policy roadmap for governments to promote immigrant inclusion, such as language access requirements and protocols for hiring practices. To raise awareness for the new program, called Certified Welcoming, the organization emailed county and city elected officials who were not already part of the organization's network. The message was sent to 18,741 representatives from 4,505 different communities.Footnote 3 Elements of the email were experimentally manipulated to isolate the effect of partisanship on policy learning.

My empirical strategy departs from previous work in two relevant ways. First, it allows me to focus on what motivates individual public officials to learn from their peers. Studies of policy diffusion have mainly focused on aggregate patterns of dissemination (Graham, Shipan and Volden Reference Graham, Shipan and Volden2013). This approach helps address several questions related to policy diffusion, but makes it difficult – if not impossible – to identify the processes leading to the adoption of a new policy. My proposed design allows me to explore previously uncovered dynamics of policy diffusion. Secondly, by collaborating with an advocacy group I am able to observe how elected officials behave in a realistic and familiar environment. Interest group contacts are only one of many ways through which officials learn about new policies. However, the pervasiveness of lobbying activities at all levels of government (de Figueiredo and Richter Reference de Figueiredo and Richter2014) gives me confidence that representatives are familiarized with this method of acquiring new information.

Experimental Conditions

The email included an endorsement by local officials from Salt Lake County, an early adopter of the policy.Footnote 4 To explore the role of partisanship in policy interest, partisan cues in the endorsement were manipulated. Since both Republicans and Democrats in Salt Lake County supported the policy, it was possible to randomly assign partisanship while holding both the policy and the endorsing municipality constant. The three versions of the endorsement are described below. Appendix A includes one version of the full email.

Local [Republicans/Democrats/Republicans and Democrats] in Salt Lake County recently joined Certified Welcoming. The program is part of an effort to attract, retain, and engage immigrants in the city.

‘Immigrants are families, homeowners, entrepreneurs, taxpayers and neighbors. Welcoming them and their contributions benefits our community’, said [Republican Lane Beattie/Democrat Ben McAdams/Lane Beattie (Rep) and Ben McAdams (Dem)] from Salt Lake County

The quote in the endorsement was taken from an op-ed article co-signed by members of both parties.Footnote 5 Local officials were randomly assigned to one of three groups: a Democratic endorsement, a Republican endorsement or an endorsement from both parties. The treatment conditions were then recoded based on the party of the local official, leading to three randomly assigned groups: (1) co-partisan endorsement for Democrats/Republicans who received a Democratic/Republican endorsement, (2) out-partisan endorsement for Democrats/Republicans who received a Republican/Democratic endorsement and (3) bipartisan endorsement.Footnote 6 Appendix C describes how officials' partisanship was measured. The treatment conditions allow me to test whether there is a co-partisan bias in policy learning.

Outcome Variables

Since I am interested in exploring public officials’ individual-level motivations to learn from the experience of others, the main outcomes in the analysis capture different levels of engagement with the certification program. I tracked who clicked on the different links available in the email. Several links were included in the message: a connection to the program's website labeled ‘Learn more about Certified Welcoming’, a link to subscribe to the organization's newsletter, and links to the organization's social media accounts. Figure A2 shows how these links were displayed.

Based on tracking information from the links, I created two complementary outcomes. Policy Interest is a binary measure that takes a value of 1 if a subject clicked on any of the links provided, and 0 otherwise.Footnote 7 To capture the level of engagement with the policy, Degree of Engagement is an ordinal variable that measures the number of links clicked by each subject. Figure D1 describes the distribution of the variable. Nearly one in four (24.8 per cent) local officials in the study clicked on two or more links. Hence, while both outcomes capture interest in learning about the policy, they measure different forms of engagement.

Results

Across the treatment conditions, 38.3 per cent of local officials clicked on one or more links in the email.Footnote 8 This result suggests that (1) the policy was deemed relevant to over a third of local officials and (2) neither ceiling nor floor effects are likely to explain the main patterns observed.Footnote 9

However, there are meaningful differences in policy interest across treatment groups. Both panels in Figure 1 describe the treatment effects of receiving an out-partisan or bipartisan endorsement, relative to a co-partisan endorsement. The outcome variables in Panels A and B are Policy Interest and Degree of Engagement, respectively. The results reveal a pattern of co-partisan bias in policy learning. Based on Panel A, on average, when the certification program is endorsed by members of the other party rather than members of the same party, officials are 4.1 percentage points less likely to demonstrate an interest in learning about it (p < 0.01). This difference represents a 10.7 per cent decrease in policy interest relative to the overall mean of the outcome.Footnote 10 The analyses based on Degree of Engagement lead to similar conclusions. On average, the number of links clicked by an individual politician decreased by 0.09 (s.e. = 0.03) when the policy was endorsed by out-partisans rather than co-partisans.Footnote 11

Figure 1.

The effects of partisan endorsements on policy learning

Note: points are estimates of the treatment effects of partisan endorsements on (a) Policy Interest and (b) Degree of Engagement. In Panel A, estimates are percentage point differences in the likelihood that an official clicked in one or more links in the email. In Panel B, estimates capture differences in the average number of links clicked. Thin/thick bars represent 95 per cent/90 per cent CIs from linear models. Table D3 reports the full model results.

These results provide an important validation of the external validity of recent studies of policy diffusion based on survey experiments (for example, Butler et al. Reference Butler2017). Partisan-based learning is not simply the result of partisan cheerleading (or satisficing) in the innocuous context of a survey (Bullock et al. Reference Bullock2013). Instead, it is reflected in behavioral efforts to learn about a new policy in what is usually an unmonitored environment. Elected officials express partisan bias in policy learning even in a context in which electoral punishment is very unlikely since their behavior was not observable by voters or competitors.

The results for the bipartisan endorsement are also consistent with a dynamic of partisan-based learning. The policy endorsed by members of both parties, on average, attracted systematically less interest than the same policy when endorsed by co-partisans. For Degree of Engagement, for instance, the difference between co-partisan and bipartisan endorsements was 0.12 (s.e. = 0.04), which represents a 17.4 per cent decrease in the number of links clicked relative to the mean value of the outcome. Moreover, the difference in levels of interest between bipartisan and out-partisan endorsements is indistinguishable from zero. For Policy Interest, the estimate for the difference in means is 0.6 percentage points (p = 0.73), while for the level of engagement (Panel B) the out-partisan endorsement generated slightly more interest than its bipartisan counterpart. The failure to reject the null hypothesis does not allow me to rule out small differences between groups. Still, the results suggest a tenuous relationship without a clear direction.

The lack of meaningful differences between out-partisan and bipartisan endorsements provides an initial explanation of why public officials may rely on partisan cues. Policies with support from both parties are perceived as compromise solutions (Harbridge, Malhotra and Harrison Reference Harbridge, Malhotra and Harrison2014). If elected officials were exclusively inferring ideological congruence from partisan cues, bipartisan endorsements should be favored over out-partisan endorsements. Appendix E shows that endorsements from both parties were also not favored over out-partisan cues in more competitive districts, where moderate policies are rewarded (Ansolabehere, Snyder and Stewart Reference Ansolabehere, Snyder and Stewart2001; Carson and Williamson Reference Carson and Williamson2018). Together, these results are consistent with a pattern of in-group favoritism. Research on in-group/out-group relations has established that animus toward out-group members spills over to ‘deviant’ in-group members who engage with the out-group (Marques, Abrams and Serôdio Reference Marques, Abrams and Serôdio2001; Stephan and Stephan Reference Stephan, Stephan and Kim2017; Tajfel and Turner Reference Tajfel and Turner1979). In this context, out-partisan and bipartisan endorsements should experience similar credibility penalties. These arguments require further investigation.

Discussion

Politicians are more likely to pursue the policies of governments that look more like them (Case, Hines and Rosen Reference Case, Hines and Rosen1993; Grossback, Nicholson-Crotty and Peterson Reference Grossback, Nicholson-Crotty and Peterson2004). However, prior work has mostly focused on aggregate patterns of policy diffusion, which makes it hard to understand the decision-making process of public officials. This study joins a recent stream of work showing how legislators’ individual characteristics drive their motivation to learn from peers.

In a field experiment that introduced politicians to a new policy endorsed by early adopters, I show that local representatives systematically demonstrate more interest in the same policy when it was previously adopted by co-partisans rather than by politicians across the aisle, or by officials from both parties. Furthermore, policies previously adopted with bipartisan support attract no more interest than out-partisan policies, even in more competitive contexts. These findings illustrate that to some degree, public officials are making decisions based on the characteristics of those who previously adopted the policy, rather than based on the content of the initiative. Shipan and Volden (Reference Shipan and Volden2012) describe similar learning processes as policy imitation: when governments copy initiatives adopted by others without concern for the policies' effects. As they explain, ‘the extent of learning in these circumstances is merely the acknowledgment that a government […] has the policy and that it must, therefore, be something desirable’ (790).

This study suffers from three main limitations. First, the external validity provided by the field experiment came with inevitable costs to the control of the intervention. Although the certification program was largely unknown, the broad policy issue of immigration is highly salient and it is reasonable to expect that local officials had strong priors regarding the value of the policy. My intuition is that partisan cues may play a larger role in less salient issues, since less information is available. Secondly, the conclusions regarding policy learning rely on the assumption that legislators who chose to ‘Learn more about Certified Welcoming’ indeed learned more about the policy. While the relationship between interest and learning has been well established (Ainley, Hidi and Berndorff Reference Ainley, Hidi and Berndorff2002), in this specific context this step was inferred. Finally, it is possible that the fact that Salt Lake County is not a standard ‘leader’ in policy innovation may have also influenced the results. Future scholarship could extend this experimental design to test whether the size of the community endorsing a policy, or its geographical proximity, impact officials' willingness to learn.

Still, the results provide relevant insights for research on policy diffusion, legislative signaling and interest group access. According to the patterns uncovered here, the prospects for policy diffusion are constrained in two meaningful ways. Partial attention to the experience of other governments may lead to inefficiencies in the policy-making process. First, public officials may be unnecessarily reinventing the wheel when only out-partisan governments have previously addressed a given policy issue. Secondly, and more problematic, in-group favoritism can lead to the persistence of poor policy solutions. If the modal policy alternative provided by co-partisans is flawed, partisan-based learning implies that poor policy choices will be reproduced.

The findings also speak to the literature on legislative signaling. Previous work suggests that heterogeneous signals (such as bipartisan endorsements) are more credible and persuasive than homogeneous signals (Krebhiel Reference Krehbiel1991). The evidence gathered here is not consistent with this argument. The certification program attracted significantly more interest when co-partisans, rather than members of both parties, endorsed it. This result is consistent with recent reassessments of information signaling models (Zelizer Reference Zelizer2018).

Finally, the study contributes to interest group research. Empirical work suggests that interest group access is mainly explained by the extent to which groups can provide resources to public officials (Eggers and Hainmueller Reference Eggers and Hainmueller2009; Kalla and Broockman Reference Kalla and Broockman2016; Salisbury Reference Salisbury1969). Yet formal theoretical work suggests that credibility plays an underappreciated role (Ainsworth Reference Ainsworth1997; Austen-Smith and Wright Reference Austen-Smith and Wright1994). The results uncovered here provide some support for the latter perspective. Future extensions of the empirical strategy adopted here can better illuminate how public officials respond to various access strategies employed by advocacy groups.

Supplementary material

Online appendices are available at https://doi.org/10.1017/S0007123420000770.

Acknowledgements

I thank Daniel Butler, Peter Loewen, David Miller, Elin Naurin, David Nickerson, Patrik Öhberg, Margit Tavits, and Adam Zelizer for the valuable feedback on the project. Early versions of this manuscript were presented at the 2018 Harvard Experimental Political Science Graduate Student Conference, the Lisbon Meeting on Political Science and Economics 2018, and the 2019 Midwest Political Science Association meeting. The study received approval from the Institutional Review Board at Washington University in St. Louis (IRB ID #201712010).

Data availability statement

Replication materials can be found in Harvard Dataverse at: https://doi.org/10.7910/DVN/SWRGQI.

Footnotes

1 Learning is only one mechanism of policy diffusion. Other mechanisms include economic competition, imitation, or coercion (Shipan and Volden Reference Shipan and Volden2008). This letter focuses on learning-based diffusion. All references to ‘diffusion’ should be interpreted as ‘learning-based diffusion’.

2 The sample is composed of elected officials from city and county governments. See Appendix B for more details.

3 The data came from two main sources: the Google Civic Information API and the American Municipal Officials Survey. Appendix B describes how this sample was created.

4 When the study was fielded, only three local governments had been certified. The risk that officials already knew about the program was limited and orthogonal to the treatments.

5 Available at http://archive.sltrib.com/article.php?id=4219261&itype=CMSID.

6 I blocked randomization on three covariates potentially related to policy interest: city/county population size, presidential vote share and data source. See Table D2 for balance tests.

7 The results are substantively similar when each category of links is analyzed separately (Table E1).

8 The analyses are restricted to public officials who opened the emails, since only these were exposed to the treatment. There were no significant differences in compliance across conditions (Table B1), and the intent-to-treat effects are substantively the same (Figure E6). Additionally, I restricted the sample of officials in the bipartisan treatment to one of the two versions of the email. The order of the parties in this condition was randomized. However, due to a coding error in one of the versions, one link was missing in the plain text version of the email. For this subgroup, the potential to engage with the policy was smaller. To ensure perfect comparability between conditions, I only consider the random sample of officials in the bipartisan condition who received the first version of the email. The results are substantively similar for the full sample (Figure E3).

9 The same is true when considering policy interest by partisanship: 40.4 per cent of Democrats and 36.7 per cent of Republicans clicked on at least one link.

10 This effect size is equivalent to the average difference in interest between officials from constituencies with Democrat and Republican majorities: 3.7 points, or 9.7 per cent of the outcome average. Survey-based estimates of co-partisan bias render similar results. Butler and Pereira (Reference Butler and Pereira2018a) find that local officials’ willingness to adopt a conservative policy is 11.5 per cent lower when it is endorsed by members of the other party.

11 As described in Figure E2, the findings are driven mainly by Democrat officials. In Appendix E, I discuss possible explanations for these patterns.

References

Abramowitz, AI and Webster, SW (2018) Negative partisanship: why Americans dislike parties but behave like rabid partisans. Political Psychology 39, 119–135.CrossRefGoogle Scholar
Adolph, C et al. (2020) Pandemic Politics: Timing State-Level Social Distancing Responses to COVID-19. Available from https://faculty.washington.edu/cadolph/papers/AABFW2020.pdf.CrossRefGoogle Scholar
Ainley, M, Hidi, S and Berndorff, D (2002) Interest, learning, and the psychological processes that mediate their relationship. Journal of Educational Psychology 94(3), 545–561.CrossRefGoogle Scholar
Ainsworth, S (1997) The role of legislators in the determination of interest group influence. Legislative Studies Quarterly 22(4), 517–533.CrossRefGoogle Scholar
Ansolabehere, S, Snyder, JM Jr and Stewart, C III. 2001. Candidate positioning in US house elections. American Journal of Political Science 45(1): 136–159.CrossRefGoogle Scholar
Austen-Smith, D and Wright, JR (1994) Counteractive lobbying. American Journal of Political Science 38(1), 25–44.CrossRefGoogle Scholar
Barabas, J and Jerit, J (2010) Are survey experiments externally valid? American Political Science Review 104(2), 226–242.Google Scholar
Barber, M and Pope, JC (2019) Does party trump ideology? Disentangling party and ideology in America. American Political Science Review 113(1), 38–54.CrossRefGoogle Scholar
Böhmelt, T et al. (2016) Party policy diffusion. American Political Science Review 110(2), 397–410.CrossRefGoogle Scholar
Bullock, J et al. (2013) Partisan bias in factual beliefs about politics. Quarterly Journal of Political Science 10(December), 519–578.CrossRefGoogle Scholar
Butler, DM and Pereira, MM (2018a) How does partisanship influence policy diffusion? Political Research Quarterly 71(4), 801–812.CrossRefGoogle Scholar
Butler, DM and Pereira, MM (2018b) Are donations to charity an effective incentive for public officials? Journal of Experimental Political Science 5(1), 68–70.CrossRefGoogle Scholar
Butler, DM et al. (2017) Ideology, learning and policy diffusion: experimental evidence. American Journal of Political Science 61(1), 37–49.CrossRefGoogle Scholar
Canes-Wrone, B, Brady, DW and Cogan, JF (2002) Out of step, out of office: electoral accountability and House members’ voting. American Political Science Review 96(1), 127–140.Google Scholar
Carson, JL and Williamson, RD (2018) Candidate ideology and electoral success in congressional elections. Public Choice 176, 175–192.CrossRefGoogle Scholar
Case, AC, Hines, JR Jr and Rosen, HS (1993) Budget spillovers and fiscal policy interdependence: evidence from the states. Journal of Public Economics 52(3): 285–307.CrossRefGoogle Scholar
de Figueiredo, J and Richter, B (2014) Advancing the empirical research on lobbying. Annual Review of Political Science 17, 163–185.CrossRefGoogle Scholar
Eggers, AC and Hainmueller, J (2009) MPs for sale? Returns to office in postwar British politics. American Political Science Review 103(4), 513–533.CrossRefGoogle Scholar
Graham, E, Shipan, CR and Volden, C (2013) The diffusion of policy diffusion research in political science. British Journal of Political Science 43(3), 673–701.CrossRefGoogle Scholar
Grossback, LJ, Nicholson-Crotty, S and Peterson, DAM (2004) Ideology and learning in policy diffusion. American Politics Research 32(5), 521–545.CrossRefGoogle Scholar
Hall, AB (2015) What happens when extremists win primaries? American Political Science Review 109(1), 18–42.CrossRefGoogle Scholar
Harbridge, L, Malhotra, N and Harrison, BF (2014) Public preferences for bipartisanship in the policymaking process. Legislative Studies Quarterly 39(3), 327–355.CrossRefGoogle Scholar
Hogan, RE (2008) Policy responsiveness and incumbent reelection in state legislatures. American Journal of Political Science 52(4), 858–873.CrossRefGoogle Scholar
Iyengar, S, Sood, G and Lelkes, Y (2012) Affect, not ideology: a social identity perspective on polarization. Public Opinion Quarterly 76(3), 405–431.CrossRefGoogle Scholar
Kalla, J and Broockman, D (2016) Campaign contributions facilitate access to congressional officials: a randomized field experiment. American Journal of Political Science 60(3), 545–558.CrossRefGoogle Scholar
Krehbiel, K (1991) Information and Legislative Organization. Ann Arbor: University of Michigan Press.CrossRefGoogle Scholar
Marques, J, Abrams, D and Serôdio, R (2001) Being better by being right: subjective group dynamics and derogation of in-group deviants when generic norms are undermined. Journal of Personality and Social Psychology 81(3), 436–447.CrossRefGoogle ScholarPubMed
Pereira, MM (2020) Replication data for: How do public officials learn about policy? a field experiment on policy diffusion, https://doi.org/10.7910/DVN/SWRGQI, Harvard Dataverse, V1, UNF:6:wCc5Tg4gysWorv4Ac7cgAw== [fileUNF]Google Scholar
Salisbury, R (1969) An exchange theory of interest groups. Midwest Journal of Political Science 13, 1–32.Google Scholar
Shipan, CR and Volden, C (2008) The mechanisms of policy diffusion. American Journal of Political Science 52(4), 840–857.CrossRefGoogle Scholar
Shipan, C and Volden, C (2012) Policy diffusion: seven lessons for scholars and practitioners. Public Administration Review 72(6), 788–796.CrossRefGoogle Scholar
Simmons, BA and Elkins, Z (2004) The globalization of liberalization: policy diffusion in the international political economy. American Political Science Review 98(1), 171–189.CrossRefGoogle Scholar
Stephan, WG and Stephan, CW (2017) Intergroup threat theory. In Kim, YY (ed.), The International Encyclopedia of Intercultural Communication. Hoboken, NJ: John Wiley & Sons, pp. 1–12.Google Scholar
Tajfel, H and Turner, JC (1979) An integrative theory of intergroup conflict. The Social Psychology of Intergroup Relations 33(47), 74.Google Scholar
Volden, C, Ting, MM and Carpenter, DP (2008) A formal model of learning and policy diffusion. American Political Science Review 102(3), 319–332.CrossRefGoogle Scholar
Zelizer, A (2018) How responsive Are legislators to policy information? Evidence from a field experiment in a state legislature. Legislative Studies Quarterly 43(4), 595–618.Google Scholar
Zelizer, A (2019) Is position-taking contagious? Evidence of cue-taking from two field experiments in a state legislature. American Political Science Review 113(2), 340–352.CrossRefGoogle Scholar
Figure 0

Figure 1.The effects of partisan endorsements on policy learningNote: points are estimates of the treatment effects of partisan endorsements on (a) Policy Interest and (b) Degree of Engagement. In Panel A, estimates are percentage point differences in the likelihood that an official clicked in one or more links in the email. In Panel B, estimates capture differences in the average number of links clicked. Thin/thick bars represent 95 per cent/90 per cent CIs from linear models. Table D3 reports the full model results.

Supplementary material: PDF

Pereira supplementary material

Pereira supplementary material

Download Pereira supplementary material(PDF)
PDF 618.8 KB
Supplementary material: Link

Pereira Dataset

Dataset

https://doi.org/10.7910/DVN/SWRGQI
Link