In the shadow of the stars and stripes: testing the malleability of U.S. support for Puerto Rican statehood

ABSTRACT Do voters update their racialized political preferences in response to new information? To answer this long-standing question, we conduct an original survey examining U.S. mainland attitudes toward towards Puerto Rican statehood, a rare consequential racialized issue of low salience. To test whether public support for statehood can be changed, we embedded an information experiment describing Puerto Rico’s political status and its relationship to the U.S. The treatment was designed to increase the perceived connection between the groups through effortful thinking. Descriptively, our results indicate that Americans are generally ambivalent to the idea of Puerto Rico becoming the 51st state. We further find that opposition to statehood is related to anti-immigration attitudes, conservative ideology, and lack of knowledge about the issue. Nonetheless, we also show that highly racialized opposition to statehood can be significantly decreased among all groups of voters by providing simple background information on U.S. and Puerto Rico’s relationship.


Introduction
Do voters update their racialized political preferences? While scholars often find that such policy preferences-defined as being linked to deeply held attitudes toward disadvantaged racial groups-are not responsive to new information, this may be a result of examining issues of high salience to voters, such as immigration. To see if this is the case, here we examine the attitudes of U.S. mainland non-Hispanic Whites toward Puerto Rican statehood-a rare consequential racialized issue which is yet of little salience to most respondents.
While Puerto Ricans have been granted U.S. citizenship since 1917, they still do not have the right to vote in federal elections and lack meaningful Congressional representation. The status question has recently regained its importance with Puerto Rico's economic crisis, which has been made worse after Hurricane Maria struck the island in 2017. In line with the result of the previous 2012 status referendum, 97% of Puerto Rican voters chose to become the 51st U.S. state in the latest 2017 election held only several months before the hurricane. 1 Any change in Puerto Rico's status, however, also requires a decision by the federal government since only the U.S. Congress can admit new states to the Union. While legislation has previously been submitted to begin the process of admitting Puerto Rico as a state, no progress has been made largely due to the limited support for the initiative in both Congress and among the U.S. public. Moreover, despite the fact that the status of Puerto Rico affects millions of U.S. citizens, there is little understanding of voter attitudes regarding the issue and whether these attitudes can be changed. On a more theoretical level, we argue that examining the malleability of attitudes toward Puerto Rican statehood can inform the growing debate on whether voters update their preferences in response to information in the case of other racialized issues such as immigration (Hopkins, Sides, and Citrin 2019;Kustov, Laaker, and Reller 2019).
To examine public attitudes toward Puerto Rican statehood, we conducted an original survey experiment with 1751 mainland non-Hispanic Whites. After completing a short back-ground survey, the respondents were randomly divided into two equal experimental groups.
While respondents in the control group were asked to read placebo information about a small tree frog native to Puerto Rico, the treatment group was exposed to information detailing Puerto Rico's historical and political connections to the United States and its current political status. The purpose of this information treatment was to increase support for Puerto Rican statehood by updating respondents' beliefs about the political relationship between Puerto Rico and the U.S., and decreasing the perceived differences between the two groups.
Overall, and in line with our expectations based on the relatively low salience of the issue, our results indicate that white Americans are generally ambivalent to the idea of Puerto Rico becoming the 51st state. Descriptively, we demonstrate greater opposition towards Puerto Rican statehood is significantly related to anti-immigration attitudes, concerns about the economy, conservative ideology, and lack of knowledge about the issue. Experimentally, we show that individual support for Puerto Rican statehood can be significantly increased by providing basic information about the historical and political connections between the U.S. and Puerto Rico. This positive effect is further present among all major demographic and political subgroups of (white) voters, including those of lower and higher initial support.
Our study makes several notable contributions. First, Puerto Rico's continuing economic turmoil and recent humanitarian crisis have brought forward many issues the island has had with governance as an unincorporated territory. One potential solution to solve Puerto Rico's issues is for the island to obtain full U.S. statehood. While our study does not aim to determine whether statehood is the socially optimal choice, our findings inform this debate by establishing that statehood could receive significant public support under certain conditions. Furthermore, we find evidence that people can be swayed to change their mind even on a significantly racialized and potentially polarizing issue such as Puerto Rican statehood, opinion about which appears to be significantly rooted in ethnocentrism and related ethnoracial predispositions. Unlike earlier work documenting the stability of immigration attitudes (ibid), our findings thus suggest that voters sometimes do update their policy preferences in light of new relevant information, at least when the issue is relatively low in salience.

Background
The question about Puerto Rico's status has been raised and debated in Puerto Rico and in the U.S. since the U.S. took control of the island from Spain in 1898. Puerto Rico is now officially considered to be a Free Associated State (Estado Libre Asociado), meaning that, while subject to federal laws like U.S. states and territories, Puerto Rico has its own constitution. Since the Jones Act of 1917, Puerto Ricans are also granted U.S. citizenship by birth. Nonetheless, while Puerto Rico residents are currently exempt from paying income taxes, they are unable to vote in federal elections or to have congressional representation.
Some argue that Congress enacted the Jones Act in order to appease Puerto Ricans and quell the calls for increased autonomy and independence. Under this view, U.S. sovereignty would reign and ensure that "Puerto Rico will never go out from under the shadow of the Stars and Stripes" (Fernandez 1992, 68). While Congress finally extended citizenship to the island with the Jones Act, it was clear that representatives were not yet willing to admit Puerto Rico as a U.S. state. Consequently, while officially an unincorporated territory of the United States, Puerto Rico's relationship with the United States has often been described as colonial in nature (ibid). For that relationship to change, not only do Puerto Ricans need to support statehood, but so does Congress. One of the major ways to persuade Congress to act may be to increase support for Puerto Rican statehood among mainland U.S. citizens.
Based on our review of existing and proposed legislation, current opposition to Puerto Rican statehood by House members mainly fall under three arguments: that Puerto Ricans are too culturally and ethnically distinct from most Americans, that the admission of Puerto Rico would be too economically costly for the U.S., and finally that Puerto Rico would skew political representation in favor of a particular party (H.R. 2499 2010). Nonetheless, there have yet to be any studies that have investigated how mainland (non-Hispanic white) Americans feel about the possibility of Puerto Rico being the 51st state, and whether these attitudes can be changed. Here we examine public opposition to Puerto Rican statehood and test its potential malleability in response to factual information about Puerto Rico.

Theoretical expectations
Given the dearth of existing research on public attitudes towards Puerto Rico and its status, we primarily rely on theoretical mechanisms previously used to explain (the change in) attitudes towards other racialized policy issues. In doing so, our study particularly builds on the vast literature examining negative U.S. public opinion on immigration, which has become increasingly linked to anti-Latino attitudes in recent years (Reny, Valenzuela, and Collingwood 2020). Although Puerto Ricans are U.S. citizens, they are Hispanic and they do migrate from the island to the mainland. Consequently, even despite the notable political differences between Puerto Ricans and other Hispanic groups on the issue of immigration specifically (Cardona-Arroyo 2020), some mainland voters may equate Puerto Rican statehood to an influx of Spanish-speaking immigrants, thereby sparking a conservative backlash against such status change (Abrajano and Hajnal 2015).
Just as members of Congress have opposed Puerto Rican statehood on economic and cultural grounds, scholars have frequently conceptualized public perceptions of immigrants as a "group threat" rooted in resource competition or concerns about cultural differences. Relatedly, research has increasingly found that public attitudes towards immigration are largely driven by voters' "sociotropic" concerns about its impact on the host society, rather than merely about their self-interest (Hainmueller and Hopkins 2014). Building on this literature, we consider the following individual-level variables as likely predictors of attitudes toward Puerto Rican statehood: partisanship and political ideology, education, immigration attitudes, economic perceptions, and knowledge of Puerto Rico. Given the observable partisan differences between Puerto Rican migrants across the country (see Cardona-Arroyo 2020) and other contextual factors, we also control for respondents' state of residence.
The important question remains: can these attitudes be changed? As growing evidence demonstrates, providing new information rarely works for immigration and related issues that are politically salient and rooted in stable ethnic or racial predispositions (see Kustov, Laaker, and Reller 2019;Tesler 2015). While being evidently racialized, Puerto Rican issues had not been widely covered by the news when our survey was administered. Therefore, it is likely that people will hesitate to come to any firm opinion regarding Puerto Rican statehood, even though their initial reaction may be shaped by their attitudes toward Latinos or more general attitudes toward marginalized ethno-racial groups. Indeed, the polls administered around the time of our survey have shown the question regarding the status of Puerto Rico has one of the highest "don't know" rates (26%). 2 However, this low issue salience also implies that most voters will lack a strong incentive to engage in motivated reasoning to defend the position of their partisan or ideological group(s) (Erisen, Lodge, and Taber 2014). Consequently, we argue that this issue provides a rare opportunity to test the malleability of racialized political preferences in response to new information.
Theoretically, our treatment aimed to increase support for Puerto Rican statehood by increasing the perceived connection between Puerto Ricans and mainland Americans. Voters have been found to oppose immigrants because they believe they are too different from themselves (Hainmueller and Hopkins 2014). By informing our respondents about the shared citizenship and experiences of Puerto Ricans and other Americans, we provide them with information that makes the connection between the two groups stronger. This connection should be further strengthened through the additional information provided in the treatment explaining that Puerto Ricans have served in the U.S. military since World War I, and that the people of Puerto Rico have defended the U.S. for more than a century. In other words, showing respondents that Puerto Ricans are willing to risk their lives to protect the U.S. is likely to shrink the perceived distinctiveness between the groups (Mandel and Litt 2013).
Given the general lack of public awareness about the issue and its low political salience, our baseline expectation is that most white Americans will neither support nor oppose the prospect of Puerto Rican statehood. Accordingly, as our main hypothesis, we also expect more white Americans to support the inclusion of Puerto Rico into the U.S. as the 51st state after being presented with background information regarding Puerto Rico's current political status. Factual information pertaining to Puerto Rico may be successful in creating support for Puerto Rican statehood because it updates respondents' beliefs about the relationship between Puerto Rico and the United States. Post-treatment, if respondents are more likely to believe that Puerto Ricans should keep their citizenship and that their current, non-state status is unfair, this suggests the treatment successfully manipulated respondents' views of Puerto Ricans as deserving of political status equal to U.S. citizens like themselves.
In addition to increasing perceived connectedness, new information about a given subject has the ability to promote "effortful thinking" (Erisen, Lodge, and Taber 2014). Requiring respondents to read background information that is directly relevant to the topic at hand encourages cognitive reasoning that is deliberate and effortful. By promoting such thinking, we are activating pathways in the brain through which individuals make semantic associations between various objects and categories. After receiving the treatment, our respondents should be directly engaging with information about Puerto Rico and comparing it to their other ideas about the island, rather than relying on predispositions to determine their policy attitudes.

Data and methods
Our study is based on a quasi-representative survey of 1751 (non-Hispanic) white U.S. mainland citizens recruited online by Cint in February 2016. This heterogeneous national sample, targeted to match census demographics, was obtained from an initial pool of 1873 respondents after accounting for response quality (attention check and survey completion). The inclusion of all respondents in the analysis, however, does not affect the results (not shown). Although younger and more educated, the sample was largely representative of the white population across important demographic and political characteristics (see Table A1). Importantly, the survey was conducted prior to Hurricane Maria and the significant political and media attention that ensued. Thus, the economic problems of Puerto Rico and the issue of its statehood were likely not on the minds of most respondents. Accordingly, we find that only 15% claimed they were very informed about Puerto Rico's debt crisis (and 37% claimed they were moderately informed). This is in line with other polls finding that few voters know that Puerto Ricans are U.S. citizens. 3 After agreeing to a standard consent form, respondents were asked various questions that sought to measure their opinion on political topics, including partisanship, ideology, perceptions of the national economy, immigration attitudes, and knowledge of Puerto Rico's debt crisis. 4 Respondents were then placed equally into either an uninformed (placebo-control) group or an informed (treatment) group. The control group was given a fact sheet about the Coqúı, a small tree frog native to Puerto Rico. In order to increase perceived connectedness of Puerto Rico through effortful thinking, the treatment group received information describing the current political status of Puerto Rico and island's historical relationship with the United States (for details, see Appendix). After being exposed to one of the treatment conditions, respondents were then asked three questions used as dependent variables (and manipulation checks). In particular, respondents were asked (1) whether it is fair for Puerto Ricans not to have any electoral votes for U.S. president; (2) whether it is fair for Puerto Ricans not to have a voting Congressional representative; (3) and whether Puerto Ricans should be allowed to keep their citizenship status. After answering these questions (combined into a "political support" index with α = 0.85), respondents received the final question used as the main dependent variable of this study: should Puerto Rico become the 51st state of the Union?

Results
Since this is the first study on this topic, we begin with a simple descriptive analysis of public attitudes toward Puerto Rico and its potential statehood. The control condition offers support for our baseline expectations that most white Americans neither support nor oppose the addition of Puerto Rico as the 51st state. The mean response in the control group as to whether Puerto Rico should become a state is 0.53 in which the plurality (32%) of respondents are undecided and the majority (69%) do not have a strong opinion either way. Still, 17% of respondents indicate that they are strongly supportive of statehood and 15% that they are strongly opposed to it. 5 While there are few differences based on demographics (including education), most factors identified by previous research as key predictors of immigration attitudes also matter for political support of Puerto Rico. Accordingly, the most supportive individuals are those who are pro-immigration, liberal, and optimistic about the national economy. There is a slight difference along party lines as well, with Democrats being more supportive of Puerto Rico than Republicans and Independents. Finally, those who report more knowledge about the debt crisis in Puerto Rico are also more supportive of statehood for the island. For details, see the regression results in Table A3.
Turning to the experimental results (Figure 1), those who received the treatment are indeed more politically supportive of Puerto Rico compared to the control group (ATE = 0.05 ± 0.02, p < 0.01). In other words, the informed participants are more likely to believe that Puerto Ricans deserve to have rights like those enjoyed by other U.S. citizens. In turn, these results suggest that, as a potential causal mechanism, our information treatment increased perceived political connections between Puerto Ricans and other Americans.
As a result, we find those in the informed group are also more likely to support Puerto Rican statehood (ATE = 0.07 ± 0.03, p < 0.01), consistent with our expectations. In particular, the informed group had significantly higher levels of support (51% vs. 40%) and lower levels of opposition (19% vs 26%) to statehood relative to the uninformed group. These effects are further robust to the inclusion of various pre-treatment covariates in the regression specification (Table A4).
Finally, we explored the possibility of heterogeneous treatment effects. While we did not pre-specify our expectations, we follow prior research and test the robustness of our results for major subgroups of respondents. Figure A1 illustrates that the treatment had a statistically similar effect regardless of respondents' partisan affiliation. Similarly, we do not find any effect differences in the case of other key moderators, including education, immigration attitudes, economic perceptions, and issue knowledge (see Table A5 and Figure A2). While the effects appear to be somewhat stronger for subgroups with lower levels of baseline support, these effect differences are not statistically significant, especially after adjusting for multiple comparisons using Bonferroni-Holm method.

Discussion and conclusion
With a population larger than in nearly half of the U.S. states, Puerto Rico has over 3 million residents who currently lack federal political representation despite their likely preference for statehood. Considering there are another approximately 5 million Puerto Ricans living in the mainland United States, the question of Puerto Rican statehood affects a sizeable share of U.S. citizens. One of the most cited reasons for the stalemate on this issue is the lack of sufficient support in the U.S. Congress and the public at large. Nonetheless, there is scant political science and public opinion research to confirm these claims.
To remedy this omission and inform policymakers, we conducted an original survey experiment. Our descriptive results show that most white Americans do not have strong views about Puerto Rico becoming the 51st state, suggesting a relatively low issue salience. Nonetheless, many voters still have meaningful, yet rather polarized, opinions on the issue. Although our particular topical focus is unique in the U.S. public opinion literature, our study demonstrates that well-established theories related to immigration and racial politics can be usefully applied to the question of Puerto Rican statehood. We find that those who oppose immigration, have negative economic perceptions, and lean conservative are more likely to oppose statehood, much like in the broader immigration politics literature.
Given that Puerto Rico is not a widely discussed topic, we also hypothesized that mainland voters (who are generally not knowledgeable about the issue) can be persuaded to support statehood by learning background information about Puerto Rico aimed at increasing its perceived connection to the U.S. mainland. The information respondents received described Puerto Ricans as having been U.S. citizens since 1917 and serving in every major U.S. war since WWI. It also explained that Puerto Ricans do not have the ability to vote for U.S. President, nor do they have a voting representative in Congress advocating for their interests. White Americans who were treated with this information were more likely to believe Puerto Ricans should maintain their U.S. citizenship, which likely caused an increase in perceptions of closeness between Puerto Rico and the United States.
The results of our experiment strongly support this interpretation. A 10% increase in support among the informed group implies that establishing a perceived connection of commonality between Puerto Ricans and white Americans is an effective way of creating support for the Puerto Rican statehood movement. In the U.S. Congress, where the decision for Puerto Rican statehood will ultimately be decided, this could be the difference that decides whether the island of 3 million achieves full U.S. statehood. It is also important to note that the treatment had a similar effect of increasing support for Puerto Rican statehood among all major subgroups of respondents, including those with lower baseline support.
While the treatment used in this experiment provides a method for overcoming the apathy and opposition felt by white Americans towards Puerto Rico becoming a state, it also demonstrates the potential of increasing perceived national connection in shaping attitudes toward other marginalized groups in the U.S. and elsewhere (also see Charnysh, Lucas, and Singh 2015). Furthermore, our results are instructive to the growing literature on various separatist and self-determination movements around the world. A new state has not been added to the U.S. since Hawaii was admitted in 1959. This is a rare phenomenon and researchers should take advantage of this opportunity to study a vibrant political movement for unification rather than secession. Finally, our findings on the malleability of attitudes toward Puerto Rican statehood inform public opinion literature on immigration that so far has rarely documented meaningful preference change in response to factual information (Kustov, Laaker, and Reller 2019).
Of course, our research is not without limitations. First, our treatment combines several lengthy and distinct pieces of information, which can all potentially have a separate effect besides their demonstrated joint effect. Such complexity, however, arguably also increases the external validity of our experiment since it is more in line with the somewhat convoluted mainstream media coverage of Puerto Rico and its status in the aftermath of Hurricane Maria. 6 Second, and related, there are uncertainties regarding whether our results would generalize to the U.S. population in the more salient post-Hurricane Maria context. Given the absence of any meaningful heterogeneous treatment effects and the limited knowledge of this subject among the general public even now, however, we are optimistic about the generalizability of our results but await future replications. Third, it is unclear to what extent an analogous treatment could increase inclusionary political attitudes in other cases where the marginalized group does not legally share a common citizenship (but see Collingwood, Lajevardi, and Oskooii 2018). Future studies could further address some of these issues by replicating our experiment for Puerto Rico and related topics in a more recent representative sample.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Supplementary material
Supplemental data for this article can be accessed https://doi.org/10.1080/ 17457289.2020.1821037.