--- title: "Information, Party Politics, and Public Support for Central Bank Independence" authors: ["Matthew DiGiuseppe", "Ana Carolina Garriga", "Andreas Kern"] year: 2025 status: "Working paper (recently accepted)" venue: "Conditionally accepted at British Journal of Political Science" url: "https://www.matthewdigiuseppe.com/papers/information-party-politics-public-support.md" links: {"PDF on SSRN": "https://papers.ssrn.com/sol3/Delivery.cfm?abstractid=5128037"} full_text: true --- # Information, Party Politics, and Public Support for Central Bank Independence DiGiuseppe, M., Garriga, A.C., & Kern, A. (2025). Information, Party Politics, and Public Support for Central Bank Independence. Conditionally accepted at British Journal of Political Science. - Status: Working paper (recently accepted) - PDF on SSRN: https://papers.ssrn.com/sol3/Delivery.cfm?abstractid=5128037 - Listed on: https://www.matthewdigiuseppe.com/#research ## Abstract Why do citizens support central bank independence (CBI)? Despite important research on economic and political reasons to grant independence to central banks, we know little about what the public thinks about CBI. This is important given citizens' potential role in constraining politicians' ability to alter CBI. We hypothesize that support for CBI is influenced by citizens' limited understanding of central bank governance and their beliefs about who will gain control over monetary policy if independence is reduced. Our expectations are confirmed by a preregistered survey experiment and a pre-post-election test in the U.S. Support for CBI increases when respondents learn that the President would gain more influence if independence was reduced. This support decreases when respondents expect a co-partisan to lead the executive branch. These findings shed light on the legitimacy basis of monetary institutions in politically polarized contexts and, from a policy perspective, indicate the limits of central bank communication. ## Full text > Extracted automatically from the preprint dated February 7, 2025; the same paper is on SSRN: https://osf.io/preprints/socarxiv/trpgz_v1. Tables, figures and equations may be garbled or missing; quote the PDF, not this text. **Information, Party Politics, and Public Support for Central Bank Independence** Matthew DiGiuseppe · Ana Carolina Garriga Leiden University* · University of Essex Andreas Kern Georgetown University February 7, 2025 Abstract Why do citizens support central bank independence (CBI)? Despite important research on economic and political reasons to grant independence to central banks, we know little about what the public thinks about CBI. This is important given citizens’ potential role constraining politicians’ ability to alter CBI. We hypothesize that support for CBI is influenced by citizens’ limited understanding of central bank governance and their beliefs about who will gain control over monetary policy if independence is reduced. Our expectations are confirmed by a preregistered survey experiment and a pre-post-election test in the U.S. Support for CBI increases when respondents learn that the President would gain more influence if independence was reduced. This support decreases when respondents expect a co-partisan to lead the executive branch. These findings shed light on the legitimacy basis of monetary institutions in politically polarized contexts and, from a policy perspective, indicate the limits of central bank communication. *Corresponding author: [email removed] Central bank independence (CBI) is under threat. Although tensions and disputes between elected officials and politically independent central banks are not unprecedented, recent efforts to exert political influence over monetary policy have reached unparalleled levels (Binder, 2021a). For example, in the UK, Tory Members of Parliament demanded a review of the Bank of England’s independence,[^1] whereas in Australia, the Greens attempted to limit the Reserve Bank’s independence.[^2] This is particularly salient in the United States, where members of the Senate have historically questioned the Federal Reserve (Fed)’s political autonomy and proposed legislation that would have automatically eliminated the Fed’s independence, such as the ‘Audit the Fed’ Bill (Binder and Spindel, 2016; Blinder, 2022). More recently, President Trump has repeatedly attacked the Fed,[^3] expressing his desire to reign in the Fed’s independence (Binder et al., 2024).[^4] Although these threats to central banks often reflect disagreement over interest rates and may be of a performative nature, these dynamics can leave the realm of rhetoric and pave the way to increasing political subordination of monetary policy. These political pressures can build the political momentum needed to reverse CBI, either formally, by changing the central banks’ statutes (Bodea and Garriga, 2023; Kern and Seddon, 2024) or de facto, by removing central bankers from office (Dreher et al., 2008). Two cases illustrate this last point. Turkey’s President Erdogan fired three consecutive central bank governors in less than two years because they refused to bend monetary policy to his unorthodox economic views.[^5] Similarly, in Hungary, the Orban administration publicly attacked Governor Simor to force his resignation. President Orban packed the Magyar Nemzeti Bank’s board of governors with loyalists, effectively reversing CBI (Adolph, 2013; Ainsley, 2017).[^6] An implicit assumption in the discussions about CBI reversals is that such a radical shift in monetary policy would enjoy popular support. However, despite growing research on general attitudes towards central banks (Anson, 2024; Blinder et al., 2024; Binder and Skinner, 2023; Cruijsen et al., 2015), there is relatively little research on the drivers of individual support for the specific policy of central bank independence. To the best of our knowledge, no research analyzes the question of whether the general public understands the concept of CBI and the political implications of reversing it. This is important for democracies where governments, next to overcoming institutional checks and balances, would require the public’s support for reforms aimed at reducing CBI. Additionally, worldwide, central banks are one of the most visible regulatory agencies that have been insulated from electoral pressures to make decisions based on technocratic grounds (Koop and Di Vettimo, 2023; Lijphart, 1984; Thatcher and Sweet, 2002). Therefore, understanding attitudes towards CBI may shed light on the drivers of public support and politicization of other non-majoritarian institutions (Thatcher et al., 2023). Building on the literature, we hypothesize that popular support for CBI is likely driven by two factors. First, citizens have limited understanding of monetary policy due to its complexity and its relative low salience (Binder, 2017; Ciuk and Yost, 2016; Coibion and Gorodnichenko, 2012; Markus, 1988). Support for CBI requires some understanding of how monetary policy works, and of how decisions are made. Because it is often unclear which political actors would gain influence over monetary policy if independence is reduced, citizens may revaluate their (lack of) support for CBI once they learn about the alternatives to CBI. Second, we expect that when respondents are informed that a decrease in CBI would lead to the executive gaining more influence, their support for CBI will be shaped by their partisan identification. Respondents will support the curtailment (increase) of CBI when co-partisans (out-partisans) are likely to control the executive power. This paper investigates individual support for CBI in an important case: the Federal Reserve during the 2024 elections. The Fed’s independence has been entrenched in law for decades. Yet, in the recent years, its independence has become object of questioning in the media and political circles. To test our hypotheses, we conducted a preregistered survey experiment using a non-probability representative sample from Prolific during the 2024 US Presidential Campaign, two weeks after Joe Biden announced he would not seek reelection. The survey timing, amid uncertainty about the race, allowed us to credibly present narratives suggesting that either candidate (Trump or Harris) could win, enhancing our ability to assess views on reducing the Fed’s independence. Our results confirm our expectations. We find that support for curtailing CBI decreases by up to 40% of a standard deviation when respondents learn that the Executive Branch will gain power if independence is reduced. Second, we find that narratives on whether a co-partisan or out-partisan is likely to win the election influences the strength of this treatment. Respondents who are told a co-partisan is likely to win the Presidency increase their support for CBI less when learning the President will gain more power. Lastly, we follow up with a portion of our original sample post-election. Consistent with our expectations, Democrats increase support for CBI while Republicans decrease support for CBI after it was clear that a Republican would be the President. We make several contributions to the literature. First, we provide a deeper understanding of what shapes public support for non-majoritarian institutions. Many agencies have been granted independence to insulate policymaking from political pressures that could lead to suboptimal decisions. By design, these agencies were not intended to seek “approval” from voters or politicians (Selin, 2015). Yet, they are increasingly contested in polarized political contexts, which puts pressure on the regulatory state (Bressman and Thompson, 2010; Koop and Lodge, 2020). Our findings highlight the importance (and limits) of providing information to citizens to enhance the legitimacy of non-majoritarian institutions (Alon-Barkat and Busuioc, 2024; Coban and Apaydin, 2024; Onoda, 2024). In doing so, our research speaks to the legitimacy of delegation in democracies (Downey, 2021; Pond, 2021) and the relationship between delegation and political polarization (Devins and Lewis, 2008; Goodhart and Lastra, 2018; Lee, 2015). Second, our study complements the fast-evolving literature on the political economy of CBI (Aklin and Kern, 2021; Baerg et al., 2021; Bodea and Hicks, 2015; Bodea and Garriga, 2023; Monnet, 2024; Moschella, 2024). Political scientists have long explored the political mechanisms underlying CBI (Adolph, 2013; Fern´andez-Albertos, 2015). Yet, the role of citizens’ attitudes towards CBI is mostly unexplored - or assumed to be embedded in partypositions (Anson, 2024; Way, 2000). In contrast with Anson (2024), we find little evidence that partisanship influence support for CBI specifically. However, the party likely to gain power in the short run plays a role in shaping opinions. Finally, our study speaks to recent research stressing potential democratic deficits of independent central banks, highlighting the need to make decision-making processes more transparent and communications clearer (Binder and Skinner, 2023; Blinder et al., 2024; Dietsch, 2020; Moschella, 2024). Our findings shows both the importance of information and suggests limitations as to how central bank transparency and communication can shape public attitudes toward CBI once they are divided across partisanship lines. Overall, our findings support the notion that, in terms of monetary policy, CBI trumps alternative arrangements and may perform as one of the institutional checks and balances on the executive branch. ### Literature Review Central bank independence denotes a structural separation between a nation’s central bank and its government (Cukierman, 1992).[^7] The key motivation for CBI is that politicians cannot credibly commit to sound monetary policymaking because they will always be tempted to (ab-)use monetary policy to win elections (Barro and Gordon, 1983; Rogoff, 1985). However, subordinating monetary policy to achieve short-term political gain comes at a cost. Electorally motivated meddling with central banks has been associated with higher inflation, undermining macroeconomic stability (Blinder, 1998; Garriga and Rodriguez, 2020, 2023). Recent evidence suggests that it can have a lasting damaging effect on central banks’ credibility, hurting a country’s international credit ratings or leading to prolonged episodes of economic turmoil (Bodea and Hicks, 2018). As a result, the key policy prescription in modern central banking has been to insulate monetary policy decisions from political interference by delegating monetary decisions to independent central banks (Barro and Gordon, 1983).[^8] A substantial literature analyzes the conditions under which governments give up control over monetary policy and increase CBI. This research, however, tends to focus on preferences over CBI for countries, governments, or political parties. However, arguments regarding economic efficiency (Bodea and Hicks, 2015; Maxfield, 1998), commitment between parties or regions (Hallerberg, 2002; Lohmann, 1998), informational asymmetries (Bernhard, 1998; Bodea, 2010; Keefer and Stasavage, 2002), diffusion of ideas among elites (Johnson, 2016; McNamara, 1998), or tying the hands of subsequent governments (Baerg et al., 2021; Boylan, 2001) do not translate directly into potential drivers of citizens’ support for CBI. Regarding individual preferences to restrict CBI, it remains unclear whether public attacks against central banks for their conduct of monetary policy reflect citizens’ opinions (Binder, 2021b; Bianchi et al., 2023).[^9] For instance, although scholars have pointed to a potential democratic deficit when monetary policy is conducted in political isolation (Hayo and Hefeker, 2002; Jones and Matthijs, 2019), we know little about what the public thinks about CBI as an institutional arrangement. Most of the research regarding individual attitudes towards central banks focuses on trust in monetary institutions.[^10] This research shows that financial literacy and general trust in institutions are associated with higher levels of trust in central banks (Brouwer and de Haan, 2022; Bursian and F¨urth, 2015; Farvaque et al., 2017; Hayo and M´eon, 2024; Kaltenthaler et al., 2010; McDowell and Steinberg, 2024).[^11] Little attention has been devoted to directly analyzing what citizens know about central banks and how this may affect support for CBI. Notable exceptions are the study by Cruijsen et al. (2015) on citizens’ knowledge of the European Central Bank, Anson (2024)’s and Blinder et al. (2024)’s analysis of the Federal Reserve, and Hayo and Neumeier (2020)’s research on the Reserve Bank of New Zealand. Examining support for CBI directly is important because citizens’ trust in monetary institutions may simply reflect government trust more broadly as a consequence the public’s ignorance of institutional configurations concerning monetary policy. Probing support for specific institutional arrangements, like CBI, can give us a better understanding of the public’s willingness to support policy change that will impact the functioning of that institution without making assumptions about the consequences of distrust or disapproval. In sum, despite recent advances in theoretical and empirical studies on the views and attitudes of citizens towards central banks, the role of citizens’ preferences for CBI remains largely unexplored or is often assumed to align with party positions (Anson, 2024; Way, 2000). To the best of our knowledge, there is no experimental evidence regarding the public’s understanding of the concept of CBI and the implications of reversing it. Given an increasing number of political attacks on central banks, it is important to address the question whether reversals of CBI would garner popular support and, if so, to understand the reasons why individuals would back such a shift in monetary policymaking. ### Argument and hypotheses #### Limited understanding of monetary institutions The starting point of our theory about preferences over CBI is the assumption of a shallow understanding of monetary policy and monetary institutions. Previous research documents the extent and effects of political ignorance (Carpini and Keeter, 1996; Gilens, 2001; Lupia, 2016), and there is evidence that people lack a good understanding of economic matters, especially when these issues have low salience (Markus, 1988; Coibion and Gorodnichenko, 2012; Ciuk and Yost, 2016; Binder, 2017). This is a key aspect for studies of individual attitudes over monetary governance: asking people about their satisfaction or trust in central banks or their preferences for CBI may produce meaningless results if respondents do not understand the questions. Building on previous research, we assume that citizens are likely unaware of the institutional alternatives to CBI – that is, who would decide monetary policy and replace the central bank if independence were reduced.[^12] Even in the context of our case, the 2024 US presidential campaign, this seems a reasonable assumption. Although both parties explicitly addressed the independence of the Fed, their calls were shallow, with no indication regarding who would fill the void if power were taken from the Fed. For example, Donald Trump repeatedly questioned the leadership of the Fed.[^13] However, he did not reveal his team’s plans to restructure the Fed until the run-up to the election in late October.[^14] Among Democrats, several legislators demanded lower interest rates.[^15] In contrast, the Biden White House issued a note stressing the salience of central bank independence.[^16] Neither of them addressed who would replace the Fed in deciding interest rates if independence was reduced. We argue that limited knowledge on this governance aspect is a key driver for support for CBI: People who are dissatisfied with the state of the economy or are just following partisan cues criticizing the central bank may want to reduce the central banks’s power and independence without reflecting on who would take charge of monetary policy decisions (instead of an independent central bank). Informing respondents of the proper counterfactual to CBI is key for our study. First, it allows us to interpret the responses as support for the central bank (Fed) being in charge of monetary policy independently (from the President) versus giving the President control over monetary policy – that is, we can interpret these answers as support for central bank independence. Without this informational treatment, it is not clear what respondents may be interpreting from the question. We expect that support for CBI will increase after receiving information about the executive power taking over monetary policy decisions for two reasons. First, trust in central banks, while low, is likely to be higher than in the executive branch (W¨alti, 2012; Ehrmann et al., 2013; Bertsou and Pastorella, 2017; Brouwer and de Haan, 2022). As such, respondents (Blinder et al., 2024; Garriga, 2025a). should prefer checks on presidential power if monetary policy decisions were subject to presidential approval or influence. Second, respondents might agree with academics that political control of interest rates is undesirable but have yet to think through the logic of reducing independence. Once presented with the alternative between decisions made by politicians or technocrats, they may be likely to prefer the latter. This is supported by research showing that citizens seem to prefer experts over politicians in technical policy domains, especially in democracies (Bertsou, 2022; Bertsou and Caramani, 2022; Chiru and Enyedi, 2022; Panel et al., 2024). In sum, we claim that ignorance over which political actors would have control over monetary policy if independence is decreased drives opposition to CBI. Therefore, we expect that informing people that the executive branch (i.e., the President) will make interest rate decisions if independence is reversed will increase public support for CBI. Consequently, our first hypothesis is: Hypothesis 1. Support for CBI will increase when respondents are provided information that the President will have greater influence over monetary policy. #### Affective polarization and support for CBI Once informed of the relationship between interest rates and economic outcomes, as well as the institutional alternatives to CBI, what factors drive support for CBI? The literature on political preferences suggests that individual preferences towards CBI could have two main sources. First, material interests affected by monetary policy decisions may play a role in support for CBI (Ansell, 2014; Bansak et al., 2021; Bearce and Tuxhorn, 2017; Scheve, 2004). Those more exposed to the effects of inflation, such as asset owners and those on fixed income, should be more inflation-averse and thus support CBI – provided that they understand the link between CBI and low inflation. Others may support reducing CBI if it implies lowering interest rates, increasing employment, and a short-run economic boom, a sentiment reflected during the US election.[^17] Individuals in the same material conditions may report different preferences depending on their risk aversion, how they weigh their egocentric and sociotropic interests, or how myopic or long-term their evaluations are.[^18] Second, ideological biases may also inform preferences for CBI (see Anson (2024)). The literature suggests that right-leaning economic preferences include lower inflation, less market intervention, and more tolerance to unemployment than left-leaning ones (Baccini and Sattler, 2023; H¨ubscher et al., 2023; Scheve, 2004). More independent central banks tend to have more conservative preferences than the government – that is, they tend to give prevalence to inflation control even in the presence of dual mandates – and are associated with fiscal restraint (Bodea and Higashijima, 2017). Therefore, holding other things constant, right-leaning economic preferences should align with support for higher CBI. Paradoxically, given dramatic shifts in partisanship platforms in the US, recent survey evidence from the US indicates that Republican voters appear to lean against CBI (Anson, 2024). Both for material and ideological interests, support for CBI would entail a preference for low inflation vis-´a-vis other economic outcomes. While we do not dispute the viability of these theoretical insights, they do not account for affective polarization (i.e., the inclination to perceive members of the opposing party unfavorably and those of one’s party favorably), a key driver in political attitudes towards institutional design. Recent evidence points out that individuals’ affective polarization drives their attitudes toward democratic norms, institutions, and policies (Graham and Svolik, 2020; Iyengar et al., 2019; Kingzette et al., 2021; Westwood et al., 2019). For instance, Graham and Svolik (2020) show that dedicated partisans back their party’s candidates even when they compromise principles like electoral fairness, institutional checks and balances, or civil liberties. Additionally, recent findings suggest that affective polarization impacts social interactions by eroding trust between groups with opposing political affiliations, shaping citizens’ political behavior (Dimant, 2024; Iyengar et al., 2019). We argue that a similar mechanism operates in the realm of monetary policy, and that affective polarization shapes citizens’ support for CBI through politicization and cue-taking (Kingzette et al., 2021). When individuals learn that reducing CBI might grant the executive branch greater influence, their support for CBI appears to be partly contingent on their partisan identification and their expectations about whether their party might control the executive branch, beyond what is explained solely by material or ideological interests. Following this line of argument, we believe that people will support the curtailment of CBI when co-partisans are likely to become President and increase their support for CBI when opposing parties are projected to win elections. Conversely, a desire to remove control from members of the opposite party will lead people to increase support for CBI when they are primed to believe the other party will control the executive branch. In the case of US politics, we expect voters identifying as Republicans to support CBI more when they anticipate the Democratic Candidate to win the presidential elections and vice versa. We synthesize these insights in our second hypothesis. Hypothesis 2. Support for CBI will decrease (increase) when respondents are provided information that the Presidency will have greater control and their party is likely to win (lose) the Presidency relative to those that are only provided information that the Presidency will have more control.[^19] ### Research Design To test our hypotheses, we fielded a preregistered survey on a non-probability sample that targeted 1,500 respondents recruited on Prolific during the US Presidential Campaign of 2024, two weeks after Joe Biden’s decision not to run for re-election.[^20] The sample reflects the national population in terms of age, sex, and partisanship.[^21] The timing of the survey and uncertainty around the state of the race at this point allowed us to test the hypotheses in one important way. The introduction of Harris to the race turned the campaign into a much more competitive race. This allowed us to provide credible narratives stating that one candidate was likely to win over the other – and potentially benefit from any reduction in the Fed’s independence. Therefore, we were able to collect data with variance on electoral expectations and potentially nudge those opinions. To test our central hypotheses, we adopted a pre-post design in which we asked about support for CBI directly before and after the informational treatment. This has two advantages. First, it allows us to examine the correlates of support for CBI before providing a treatment. Second, it increases the power of the experiment, although with a small risk of inducing consistency bias (Clifford et al., 2021). However, such bias should decrease the size of any effects. Before the treatments and collection of the outcome variable(s), we provided a brief explanation of how interest rates are set and the Fed’s current independence to anchor respondents. It reads as follows: “You may have noticed that interest rates have also been unusually high for the past several years. When interest rates are higher, loans for things like buying a house or starting a business cost more because the interest you have to pay back is higher. For people or businesses already paying back loans, it can mean their payments go up if their loans and credit cards have variable or adjustable interest rates. On the flip side, when interest rates increase, saving money becomes more attractive. Banks offer higher interest on savings accounts or certificates of deposit, so the money you have saved grows faster. Why do interest rates fall and rise? Interest rates change because of the decisions of the Federal Reserve, often called “the Fed”. One of the Fed’s major goals is to reduce inflation. To reduce inflation, the Fed has to increase interest rates. Increasing interest rates makes it more expensive to borrow and make investments. This has the effect of slowing down the economy and reducing inflation. Unfortunately, this usually means more unemployment and lower wages. This happens because businesses make fewer investments, and people buy fewer things. For example, with higher interest rates, it costs more to build and buy a house. This means fewer home builders will be hired. While unfortunate, this is usually necessary to bring down inflation. The Federal Reserve functions as an independent entity within the framework of the U.S. government without the direct intervention of elected officials. This independence is achieved through a carefully structured appointment process for its members and its ability to generate its own funding. The Board of Governors of the Federal Reserve, which includes its Chairperson, is appointed by the President and must be confirmed by the Senate. These officials serve staggered terms that usually last longer than the tenure of any single President, ensuring a level of continuity and stability in monetary policy. Additionally, it is difficult to remove Fed officials before their terms expire, except in cases of ethical or legal misconduct. This arrangement grants them considerable independence to act without considering the political impact of unpopular decisions.” We then ask: “People disagree about the independence of the Federal Reserve. Do you think this independence should be decreased or increased?” Respondents answered on a 5-point scale from ‘decreased a lot’ to ‘increased a lot.’ The answer to this question serves as our pre-treatment indicator of CBI. In the Supplementary Appendix we also use it as a dependent variable to examine the correlates of CBI before we treat with information about its counterfactual. Following this question, we provide a short statement about differing perspectives of the Fed’s history to respondents in each treatment condition. This primarily serves to avoid asking the pre- and post-treatment questions back-to-back in the control group.[^22] After the assignment of the treatment, we then ask respondents to again indicate their support for CBI: “We will ask again, do you think this independence of the Federal Reserve should be decreased or increased?” The answer to this question is our primary dependent variable. Figure 1 plots the distribution of both the pre-and post-treatment responses. At least in our sample, close to a majority of respondents prefer the status quo, and about 30% of respondents wish to see independence reduced, while less than 20% of respondents are interested in increasing independence. As panel B of Figure 1 illustrates, we see little variation in support for CBI across party identification. #### Treatment Arms Our experiment assigns four treatment arms via block randomization based on party affiliation (Democrat: 0-2, Republican: 4-6, and Independent: 3 on the 7-point party identification Figure 1: Distributions of Responses on CBI Independence: Panel A shows the distribution of support for CBI pre- and post-treatment for all observations. Panel B shows the mean support for pre-treatment CBI for respondents in the control group (N=366). scale), with equal probability within blocks. The control arm simply asks respondents to restate their answer to the question: “We will ask again, do you think this independence of the Federal Reserve should be decreased or increased?” The first treatment arm simply notifies respondents that the President will likely gain more power over interest rates if CBI is decreased AND that the outcome of the election is very uncertain: “If the Federal Reserve (the Fed) had less independence, more power would likely be given to the President. Future Presidents could be given more power to overturn Fed policy and remove Fed officials, or in extreme cases; the executive branch would set interest rates themselves.” The second treatment arm treats with the same information but adds that Republicans have an advantage in winning the Presidential race: “If you’ve looked at the polls recently, you’ll see that Former President Trump is ahead and has a strong chance of becoming president again. Betting markets give him a 65% probability of winning. While Harris has narrowed the gap in the popular vote, Trump still commands a sizeable lead in key swing states and is very likely to win in the electoral college and thus win the Presidency.” N Congress President Treasury Not Mentioned ```text Control 352 9.9 12.3 13.7 64.0 Pooled Treatment 1041 4.2 75.1 7.2 13.5 ``` Table 1: Attention Check: This table indicates the percent of respondents that recalled information regarding which actor gained more influence if CBI was decreased in either the pooled treatment or control conditions. We asked all respondents the following question: “Above we told you who would take more responsibility for interest rates if the Fed’s independence was reduced. Who did we mention would have more control on interest rates?”. The options are listed in the table. The third treatment arm also provides information about the President but instead attempts to increase the perception that Democrats are likely to win: “If you’ve looked at the polls recently, you’ll see that Democrats are doing very well with key swing state demographics after Kamala Harris took over as presumptive nominee. Many experts now think Democrats will retain the White House. Recent high-quality polls show Kamala Harris is leading Donald Trump by 2-points Nationally. Other polls show she has taken a lead in must-win swing states like Pennsylvania.” Our final two treatments differ because we wanted to connect the expectations of winning with realistic justifications. Each candidate, at the time, had specific paths to victory. Treating otherwise risked reducing the strength of the treatment.[^23] Table 1 presents the results of an attention check aimed to capture if respondents recall being told the President would gain more control in the event of reduced CBI. Over 75% of those assigned to one of the three treatments could recall the information later in the survey. 64% correctly noted that we did not mention a counterfactual. #### Covariates and Electoral Expectations We collected several variables that corresponded with the expectations in the literature regarding the relationship between information, partisanship, and CBI. We use these covariates in our experimental models to increase the precision of our estimates. As mentioned above, we also use these covariates to examine the correlates of pre-treatment opinions on CBI (see Supplementary Appendix). The covariates include party identification, age, income, college education, mortgage holding, inflation impact on individual welfare, a count of loans and assets, responsibility for household financial decisions and grocery shopping, financial literacy, knowledge of the Federal Reserve, and trust in various government branches and institutions (including the Fed). We provide more detail about these variables in the Supplementary Appendix. In general, these covariates are not strong predictors of baseline CBI attitudes. In several linear models estimating support for CBI (presented in the Supplementary Appendix), with over 10 covariates, we find an R[^2] no larger than 0.14. Notably, party identification and income do not have a strong relationship with support for CBI.[^24] Instead, we find that those who report to have suffered from inflation are more supportive of CBI. Consistent with our expectations, those who have more exposure to finance via financial decision-making at home or a college education – likely proxies for more information regarding monetary issues – are more supportive of CBI. Beyond that, however, financial literacy and knowledge about the Fed are not strong predictors of CBI support. Lastly, general trust in the Fed has a positive association with CBI support. Yet, the substantive correlation is small. Despite their weak relationship with support, these covariates may add precision to our estimate of the treatment effects. As such, we include them on the right-hand side of a linear model that estimates the effect of the treatment. This approach increases statistical power and researcher degrees of freedom. Consequently, we select these variables agnostically with a LASSO selection model following the recommendation of Bloniarz et al. (2016). We include each continuous variable individually and each categorical variable as dummies in a model predicting the outcome. The LASSO model returns only variables with non-zero coefficients that are then included in our estimation of the treatment effects in addition to the pre-treatment CBI measure. ### Results Our first pre-registered hypothesis suggests that information about the actor that gains power in the event of reducing CBI (the President) should increase support for CBI. Figure 2 presents the average treatment effects (ATE) of each treatment arm and the pooled treatment arms. In each estimation of the ATE, we standardize the dependent variable (mean=0, SD=1) and present robust standard errors. The information treatment omitting electoral expectations increases support for CBI by 24% of a standard deviation.[^25] For full transparency, the figure shows the ATE for the pooled information treatment and each of the treatment arms. Additionally, Figure 2 also shows the complier average causal effect (CACE). This estimate reflects the coefficient of a two-stage least squares estimation in which the treatment is an instrument for whether or not the respondent could recall the information later in the survey (see Table 1). Therefore, it presents the effect among those who were assigned the information treatment and could recall the information. Unsurprisingly, once we account for inattentive respondents, the effect size increases to 41% of a standard deviation increase in the support for CBI scale, albeit with a larger standard error. This provides strong support for the first hypothesis: Support for CBI increases when respondents are informed of the counterfactual that the President would gain more power. Figure 2: Information Effects: This figure shows the average treatment effect (ATE) and complier average causal effect (CACE) of informing respondents that reducing central bank independence likely means that the President will gain greater power. Two of the treatment arms also intend to treat expectations that either a Democrat or Republican will win the Presidency. The top 3 coefficients show the ATE of each individual arm (blue). The red dot shows the ATE of the pooled treatments compared to the control condition. The final coefficient indicates the CACE of the pooled treatment. Each point estimate indicates a standardized coefficient and the bars indicate the 95% confidence intervals. Next, we examine the effect of the electoral expectation treatments. Figure 3 presents two panels. Each plots the ATE for those who received a message that a co-partisan is likely to win the election and for those who received a message that an out-partisan will win. The estimates in each panel reflect different baselines - the control group on the left and the information treatment without expectations on the left. Importantly, we only include respondents who identify with a political party in this analysis, excluding ‘leaners’ and independents. We then recode the treatments to pool those who received a co-party victory treatment or those who received an out-party victory treatment. The left panel is informative because it shows that regardless of how we framed the outcome of the election, the counterfactual increases support for CBI among partisans.[^26] Yet, the effect is stronger among those who are informed that an out-party is likely to win the Presidential election. Those informed that the president will take control and an out-party is likely to win increase their support by over 44% of a S.D. compared to the control that did not receive information about the president or election. Learning a the president will gain greater interest rate power and a narrative of co-party victory only increase support for CBI by 15% of a S.D. Figure 3: CATE Election Expectations: The two panels in this figure show the Conditional Average Treatment Effect of informing respondents that either a member of their party or a member of the opposing party is likely to win the presidential election. The left panel compares these treatments against the control group that received no information (N=815). The right panel compares these treatments against the information treatment group that did not receive electoral expectations (N=897). The latter reflects our preregistered analysis. Each point estimate indicates a standardized coefficient and the bars indicate the 95% confidence intervals. The right panel of Figure 3, our preregistered comparison, shows that compared to the pure information treatment, those who receive information that a co-party candidate will win decrease their support for CBI by 13% [-28%, -2% ] of a standard deviation. Conversely, we see that the out-party treatment does increase support for CBI but it is insignificant at our preregistered alpha value 0.95. The small effect sizes are potentially a consequence of the difficulty of changing expectations of the state of the presidential race. At this time, the respondents we are polling are likely responding to many surveys about the presidential race and are also immersed in the broader information environment. This aligns with our efforts to record a manipulation check. We find that the co-party treatment did influence expectation of co-party victory relative the control, but it was not significant relative to those in the pure information treatment condition. However, there was very little variance in co-partisan electoral expectations. A strong majority (80%) expected a co-partisan to win the election. #### Results by Party The results above may be a product of either Democrats or Republicans having different expectations of co-partisan victory and different responses to information. As such, we might have lower confidence that the effects hold broadly and are rather just present among a specific party (Anson, 2024). With the caveat of lower statistical power, we also examine the partisan differences in the treatment effects. Figure 4 presents the CATE for those who identify as Republicans or Democrats. The information treatments are statistically significant for all treatment-sample pairs except Democrats who also received a co-partisan election treatment. This gives us greater confidence that information can play an important role in shaping attitudes across the political spectrum. We find, however, that the electoral expectation treatment effect can be primarily attributed to Democratic respondents. Although we can only speculate about the reasons for this difference, one explanation is that it was more difficult to alter the beliefs of Republicans regarding the election outcome, as Trump had a clear lead until Harris came into the race. However, this interpretation also aligns with pessimistic views among Democratic voters about a potential Trump administration attempting to bend the boundaries of democratic governance (Voelkel et al., 2024). Figure 4: Effects by Party: The top panel shows the ATE for each treatment arm, compared to the control arm, on Democratic and Republican subsets of the main sample. The bottom two panels show the effects of the electoral expectations treatment arms compared to the information treatment arm for Democrats (left) and Republicans (right). #### Material Interests or Polarization? The evidence above suggests clear partisan differences in their response to the informational treatments. However, there is still little evidence to distinguish whether these differences are driven by preferences over monetary policy (low vs. high interest rates) or by affective polarization. A deeper look into our data suggests that partisan differences regarding CBI are not driven by material interests. First, we asked respondents to indicate if they would personally benefit from high or low interest rates on a 3-point, 0-2, scale (high, indifferent, low) before the treatments.[^27] In Figure 5, we plot the mean responses to this question across party identification. We see very little difference in the outcome. This finding strongly suggests that previously observed differences in preferences over interest rate policy are not found in our individual-level data (Quinn and Shapiro, 1991; Mukherjee and Leblang, 2007; Clark and Arel-Bundock, 2013). Figure 5: Mean Interest Rate Preferences by Party ID: Each bar shows the mean preferences for low interest rates at each level of the 7-point party identification scale (Strong Democrat - Strong Republican). Interest rate preferences are coded as following - 0 =”Interest rates go up (increasing the return on my savings)”, 1 = ”Interest rates stay the same” or ”I don’t know”, 2 = ”Interest rates go down (making credit card interest lower and loans cheaper)” Next, we show that the effect of the treatment does not significantly vary across preferences for interest rate policy (the way it does across co-party and out-party control of the presidency). If policy preferences were driving the conditionality of the treatment effect, we would expect the treatment effect to be larger for those who prefer higher interest rates (at the lower end of the scale). As Figure 6 shows, there is no statistical difference across the scale, and the substantive difference is small. A follow-up analysis that estimates the model for separately by partisanship indicates that this non-finding holds for members of each party. ### Pre-Post Election Analysis Following the election, we re-contacted 1,022 of the original 1,500 respondents (68%) of our survey on the Prolific platform to examine how the election itself changed CBI opinions.[^28] Figure 6: Marginal Effect of Pooled Treatment Across Interest Rates preferences: The figure shows the CATE of the pooled information treatment arms across pre-treatment measures of interest rate preferences (high to low). The black line represents the marginal effect and the ribbon indicates the 95% confidence intervals. We expect that having Republicans won the presidency, Democrats would increase their support for CBI and Republicans would decrease their support for CBI.[^29] This exercise serves as an important validation of our initial expectation that partisanship alignment between the executive and the respondents plays a key role in shaping attitudes toward CBI. To recover the impact of the election of support for CBI, we first estimate a linear model in which we include observations from both the first and second waves of our survey. The dependent variable is the post-treatment CBI measure for either the first or second wave of the survey. We then either include an interaction between the wave indicator and the 7-point party identification measure or we estimate the effect on separate samples of Republicans and Democrats. In each model, we include respondent fixed-effects and robust standard errors. Figure 7 plots the marginal “effect” of the election across partisanship (Panel A) and for members of each party (Panel B). The results strongly align with our theoretical expectations. Democrats indicate stronger support for CBI. Republicans, who now have a co-partisan President elect, demonstrate lower support for CBI post-election. Panel C plots the movement of respondents across the likert scale. This plot helps us further understand Figure 7: Post-Election Shifts: Panel A shows the marginal effect of the election on attitudes toward CBI. The line indicates the estimate, and the shaded areas indicate the 95% confidence intervals. Panel B plots the effect of the election on separate samples of Republican and Democrats. The bars indicate the 95% confidence intervals. Panel C presents 2 Sankey plots indicating the shifts in responses for Democratic and Republican voters respectively between survey waves. A 1 indicates strong support for reducing CBI. A 5 indicates strong support for increasing CBI. if the results are driven by one-party or by movements on one pole of the scale. Notably, we see that among Democrats, a non-trivial number of respondents who were in favor of reducing CBI now wish to keep CBI untouched. Even more Democrats who wanted no changes to CBI, now favor increasing CBI (3 to 4 or 5 on the scale). In general, the proportion of flows seem to be consistent across values suggesting a more general shift in opinion. On the Republican side, we see a sizable retreat from those wishing to maintain independence, now preferring to reduce CBI. We see relatively few respondents on the Republican side increasing their support for independence suggesting that, despite the unpopularity of Trump among some Republicans, there is little indication of ‘never Trumpers’ looking to restrain executive power. Again, the proportion of movements is consistent across the scale. It is worth noting that the effect of the election is not causally identified in a strict sense. Many things happened in the 3+ months between the survey waves that could have changed attitudes about CBI. However, we are hard-pressed to think of any events that could explain a conditional change in attitudes toward CBI besides the election result.[^30] ### Conclusions Central bank independence (CBI) has been the gold-standard institutional arrangement for central banking. Built on the premise that monetary policy should not be left to politicians, the literature has emphasized the political and economic benefits of CBI. However, CBI also creates an accountability imbalance and sets the stage for conflict between elected officials and independent central bankers (Binder, 2021a; Bodea and Garriga, 2023). These conflicts tend to surface when elected officials, unable to use the money printing press to reactivate the economy, or to curb political backlash in times of inflation or high interest rates, direct criticism at the central bank’s leadership. Although such political efforts often represent mere virtue signaling or blame deflection, verbal attacks and pressure on central banks could generate momentum to challenge CBI. Whether the public would support such a political rollback of CBI remains unclear. Yet, public support is key for important policy changes in democratic contexts. Our case study is also particularly important. Given the current political climate in the United States, where the Fed faces increasing political opposition, understanding public support for CBI is timely and can shed light to support for reforms or policies affecting other longstanding institutions. Assuming a shallow understanding of monetary policy, and drawing on a growing literature on political polarization (Graham and Svolik, 2020; Iyengar et al., 2019; Kingzette et al., 2021), we argued that citizens’ support for CBI is influenced both by information about monetary policy and by partisan identification – the later becomes salient when individuals learn that reducing CBI would increase the President’s influence. We predicted that respondents will favor curtailing CBI when their preferred party is likely to hold the presidency and will support maintaining or increasing CBI if an opposing party is expected to take office. To test our theoretical predictions, we conducted a pre-registered survey experiment during the 2024 U.S. Presidential campaign, shortly after Biden announced that he would not seek re-election. The survey timing allowed us to credibly present scenarios suggesting that either party could win, helping us gauge public attitudes toward reducing the Fed’s independence. We find that voters from across the political spectrum increase their support for CBI when they are informed that the President will likely gain influence if CBI is decreased. However, we also find that informing respondents about the likelihood of a co-partisan winning office conditions this response. Following the election, we find strong changes in support for CBI among the same respondents. Democrats increase their support and Republicans decrease their support. Our findings have several important implications. First, by analyzing support for CBI, we provide new insights into the public’s understanding of monetary policy and support for independent central banks. Our results indicate that central bank communications and greater transparency are key to promote support for CBI. However, information may have limited firepower to fend off political pressure to reverse CBI when political polarization is pronounced. Second, when Presidents enjoy widespread support, citizens aligned with the President’s party are more inclined to support curtailing CBI, even if it does not serve their long-term self-interest. This effect is independent of partisan affiliation and material interests, indicating potential vulnerabilities for central banks without deep institutional anchoring. While we recognize the value of central bank accountability, our findings indicate that in an increasingly polarized world, transparency may not be an effective way to increase support for independence in the eyes of the public, who may acquiesce to political interference if co-partisans benefit in the short run, threatening a central bank’s political independence. Overall, our findings support the notion that CBI constitutes an institutional arrangement that provides additional checks and balances on the executive branch of government, and it may be seen as such by voters who would prefer less concentration of political power. As such, independent central banks also constitute a vital pillar of democratic governance, balancing competing interests and safeguarding macro-financial stability in times of deep political divisions. ### References Adolph, Christopher (2013). Bankers, bureaucrats, and central bank politics: The myth of neutrality. Cambridge University Press. Adrian, Tobias , Ashraf Khan, and Lev Menand (2024). A new measure of central bank independence. IMF Working Paper No. 24/35. Ainsley, Caitlin (2017). The politics of central bank appointments. The Journal of Politics 79 (4), 1205–1219. Aklin, Micha¨el and Andreas Kern (2021). The side effects of central bank independence. American Journal of Political Science 65 (4), 971–987. Alon-Barkat, Saar and Madalina Busuioc (2024). Independent agencies, credible policies? the role of prior beliefs in shaping stakeholder perceptions of credibility in a contested environment. Journal of European Public Policy , 1–32. Ansell, Ben (2014). The political economy of ownership: Housing markets and the welfare state. American political science review 108 (2), 383–402. Anson, Ian G (2024). Fed up: The determinants of public opposition to the us federal reserve. Political Research Quarterly , 10659129241265703. Baccini, Leonardo and Thomas Sattler (2023). Austerity, economic vulnerability, and populism. American Journal of Political Science. Baerg, Nicole Rae , Julia Gray, and Jakob Willisch (2021). Opportunistic, not optimal delegation: The political origins of central bank independence. Comparative Political Studies 54 (6), 956–988. Bansak, Kirk , Michael M Bechtel, and Yotam Margalit (2021). Why austerity? the mass politics of a contested policy. American Political Science Review 115 (2), 486–505. Barro, Robert J and David B Gordon (1983). Rules, discretion and reputation in a model of monetary policy. Journal of monetary economics 12 (1), 101–121. Bearce, David H and Kim-Lee Tuxhorn (2017). When are monetary policy preferences egocentric? evidence from american surveys and an experiment. American Journal of Political Science 61 (1), 178–193. Bernanke, Ben S (2022). 21st century monetary policy: The Federal Reserve from the great inflation to COVID-19. WW Norton & Company. Bernhard, William (1998). A political explanation of variations in central bank independence. American Political Science Review 92 (2), 311–327. Bertsou, Eri (2022). Bring in the experts? citizen preferences for independent experts in political decision-making processes. European Journal of Political Research 61 (1), 255– 267. Bertsou, Eri and Daniele Caramani (2022). People haven’t had enough of experts: Technocratic attitudes among citizens in nine european democracies. American Journal of Political Science 66 (1), 5–23. Bertsou, Eri and Giulia Pastorella (2017). Technocratic attitudes: A citizens’ perspective of expert decision-making. West European Politics 40 (2), 430–458. Bianchi, Francesco , Roberto G´omez-Cram, Thilo Kind, and Howard Kung (2023). Threats to central bank independence: High-frequency identification with twitter. Journal of Monetary Economics 135, 37–54. Binder, Carola C (2017). Fed speak on main street: Central bank communication and household expectations. Journal of Macroeconomics 52, 238–251. Binder, Carola C (2021a). Political pressure on central banks. Journal of Money, Credit and Banking 53 (4), 715–744. Binder, Carola C (2021b). Presidential antagonism and central bank credibility. Economics & Politics 33 (2), 244–263. Binder, Carola C and Christina P Skinner (2023). The legitimacy of the federal reserve. Stan. JL Bus. & Fin. 28, 1. Binder, Sarah , Jackson Gode, Christian Potter, and Mark Spindel (2024). Tweeter-in-chief: Dynamics of trump’s populist campaign against the fed. Presidential Studies Quarterly . Binder, Sarah and Mark Spindel (2016). Congress and the federal reserve. In Congress and Policy Making in the 21st Century. Cambridge University Press. Blanchflower, David G and Conall MacCoille (2009). The formation of inflation expectations: an empirical analysis for the uk. Technical report, National Bureau of Economic Research. Blinder, Alan S. (1998). Central Bank in Theory and Practice. Cambridge, MA: MIT Press. Blinder, Alan S (2022). A Monetary and Fiscal History of the United States, 1961–2021. Princeton University Press. Blinder, Alan S , Michael Ehrmann, Jakob De Haan, and David-Jan Jansen (2024). Central bank communication with the general public: Promise or false hope? Journal of Economic Literature 62 (2), 425–457. Bloniarz, Adam , Hanzhong Liu, Cun-Hui Zhang, Jasjeet S Sekhon, and Bin Yu (2016). Lasso adjustments of treatment effect estimates in randomized experiments. Proceedings of the National Academy of Sciences 113 (27), 7383–7390. Bodea, Cristina (2010). Exchange rate regimes and independent central banks: a correlated choice of imperfectly credible institutions. International Organization 64 (3), 411–442. Bodea, Cristina and Ana Carolina Garriga (2023). Central bank independence in latin america: Politicization and de-delegation. Governance 36 (1), 59–80. Bodea, Cristina and Raymond Hicks (2015). International finance and central bank independence: Institutional diffusion and the flow and cost of capital. The Journal of Politics 77 (1), 268–284. Bodea, Cristina and Raymond Hicks (2018). Sovereign credit ratings and central banks: Why do analysts pay attention to institutions? Economics & Politics 30 (3), 340–365. Bodea, Cristina and Masaaki Higashijima (2017). Central bank independence and fiscal policy: Can the central bank restrain deficit spending? British Journal of Political Science 47 (1), 47–70. Boylan, Delia M (2001). Defusing democracy: Central bank autonomy and the transition from authoritarian rule. University of Michigan Press. Bressman, Lisa Schultz and Robert B Thompson (2010). The future of agency independence. Vand. L. Rev. 63, 599. Brouwer, Nils and Jakob de Haan (2022). Trust in the ecb: Drivers and consequences. European Journal of Political Economy 74, 102262. Bursian, Dirk and Sven F¨urth (2015). Trust me! i am a european central banker. Journal of Money, Credit and Banking 47 (8), 1503–1530. Carpini, Michael X Delli and Scott Keeter (1996). What Americans know about politics and why it matters. Yale University Press. Chiru, Mihail and Zsolt Enyedi (2022). Who wants technocrats? a comparative study of citizen attitudes in nine young and consolidated democracies. The British Journal of Politics and International Relations 24 (1), 95–112. Ciuk, David J and Berwood A Yost (2016). The effects of issue salience, elite influence, and policy content on public opinion. Political Communication 33 (2), 328–345. Clark, William Roberts and Vincent Arel-Bundock (2013). Independent but not indifferent: Partisan bias in monetary policy at the fed. Economics & Politics 25 (1), 1–26. Clifford, Scott , Geoffrey Sheagley, and Spencer Piston (2021). Increasing precision without altering treatment effects: Repeated measures designs in survey experiments. American Political Science Review 115 (3), 1048–1065. Coban, M Kerem and Fulya Apaydin (2024). Navigating financial cycles: Economic growth, bureaucratic autonomy, and regulatory governance in emerging markets. Regulation & Governance. Coibion, Olivier and Yuriy Gorodnichenko (2012). What can survey forecasts tell us about information rigidities? Journal of Political Economy 120 (1), 116–159. Cruijsen, Carin , David-Jan Jansen, and Jakob de Haan (2015). How much does the public know about the ecb’s monetary policy? evidence from a survey of dutch households. International Journal of Central Banking 11 (4), 169–218. Cukierman, Alex (1992). Central Bank Strategy, Credibility, and Independence: Theory and Evidence. MIT Press. Devins, Neal and David E Lewis (2008). Not-so independent agencies: Party polarization and the limits of institutional design. BUL rev. 88, 459. Dietsch, Peter (2020). Independent agencies, distribution, and legitimacy: The case of central banks. American political science review 114 (2), 591–595. Dimant, Eugen (2024). Hate trumps love: The impact of political polarization on social preferences. Management Science 70 (1), 1–31. Downey, Leah (2021). Delegation in democracy: A temporal analysis. Journal of Political Philosophy 29 (3), 305–329. Dreher, Axel , Jan-Egbert Sturm, and Jakob De Haan (2008). Does high inflation cause central bankers to lose their job? evidence based on a new data set. European Journal of Political Economy 24 (4), 778–787. D’Acunto, Francesco , Evangelos Charalambakis, Dimitris Georgarakos, Geoff Kenny, Justus Meyer, and Michael Weber (2024). Household inflation expectations: An overview of recent insights for monetary policy. NBER Working Paper No. 32488. Ehrmann, Michael , Michel Soudan, and Livio Stracca (2013). Explaining european union citizens’ trust in the european central bank in normal and crisis times. The Scandinavian Journal of Economics 115 (3), 781–807. Farvaque, Etienne , Muhammad Azmat Hayat, and Alexander Mihailov (2017). Who supports the ecb? evidence from eurobarometer survey data. The World Economy 40 (4), 654–677. Fern´andez-Albertos, Jos´e (2015). The politics of central bank independence. Annual Review of Political Science 18, 217–237. Garriga, Ana Carolina (2025a). Public confidence in the bank of england. National Institute Economic Review , 1–23. Garriga, Ana Carolina (2025b). Revisiting central bank independence in the world: An extended dataset. International Studies Quarterly (Forthcoming). Garriga, Ana Carolina and Cesar M Rodriguez (2020). More effective than we thought: Central bank independence and inflation in developing countries. Economic Modelling 85, 87–105. Garriga, Ana Carolina and Cesar M Rodriguez (2023). Central bank independence and inflation volatility in developing countries. Economic Analysis and Policy 78, 1320–1341. Gilens, Martin (2001). Political ignorance and collective policy preferences. American Political Science Review 95 (2), 379–396. Goodhart, Charles and Rosa Lastra (2018). Populism and central bank independence. Open Economies Review 29, 49–68. Graham, Matthew H and Milan W Svolik (2020). Democracy in america? partisanship, polarization, and the robustness of support for democracy in the united states. American Political Science Review 114 (2), 392–409. Hallerberg, Mark (2002). Veto players and the choice of monetary institutions. International Organization 56 (4), 775–802. Hayo, Bernd and Carsten Hefeker (2002). Reconsidering central bank independence. European Journal of Political Economy 18 (4), 653–674. Hayo, Bernd and Pierre-Guillaume M´eon (2024). Preaching to the agnostic: Inflation reporting can increase trust in the central bank but only among people with weak priors. Journal of International Money and Finance 142, 103037. Hayo, Bernd and Florian Neumeier (2020). Public knowledge about and attitudes towards central bank independence in new zealand. Journal of Banking & Finance 113, 105737. H¨ubscher, Evelyne , Thomas Sattler, and Markus Wagner (2023). Does austerity cause polarization? British Journal of Political Science 53 (4), 1170–1188. Iyengar, Shanto , Yphtach Lelkes, Matthew Levendusky, Neil Malhotra, and Sean J Westwood (2019). The origins and consequences of affective polarization in the united states. Annual review of political science 22 (1), 129–146. Johnson, Juliet (2016). Priests of Prosperity: How Central Bankers Transformed the Postcommunist World. Ithaca: Cornell Univeristy Press. Jones, Erik and Matthias Matthijs (2019). Rethinking central-bank independence. Journal of Democracy 30 (2), 127–141. Kaltenthaler, Karl , Christopher J Anderson, and William J Miller (2010). Accountability and independent central banks: Europeans and distrust of the european central bank. JCMS: Journal of Common Market Studies 48 (5), 1261–1281. Keefer, Philip and David Stasavage (2002). Checks and balances, private information, and the credibility of monetary commitments. International organization 56 (4), 751–774. Kern, Andreas , Bernhard Reinsberg, and Matthias Rau-G¨ohring (2019). Imf conditionality and central bank independence. European Journal of Political Economy 59, 212–229. Kern, Andreas and Jack Seddon (2024). The politics of reversing central bank independence. Journal of Political Institutions and Political Economy 5 (3), 443–485. Kingzette, Jon , James N Druckman, Samara Klar, Yanna Krupnikov, Matthew Levendusky, and John Barry Ryan (2021). How affective polarization undermines support for democratic norms. Public Opinion Quarterly 85 (2), 663–677. Koop, Christel and Michele Scotto Di Vettimo (2023). How do the media scrutinise central banking? evidence from the bank of england. European Journal of Political Economy 77, 102296. Koop, Christel and Martin Lodge (2020). British economic regulators in an age of politicisation: From the responsible to the responsive regulatory state? Journal of European Public Policy 27 (11), 1612–1635. Lee, Frances E (2015). How party polarization affects governance. Annual review of political science 18 (1), 261–282. Lijphart, Arend (1984). Democracies: Patterns of Majoritarian and Consensus Government in Twenty-One Countries. New Haven: Yale University Press. Lohmann, Susanne (1998). Federalism and central bank independence: the politics of german monetary policy, 1957–92. World Politics 50 (3), 401–446. Lupia, Arthur (2016). Uninformed: Why people know so little about politics and what we can do about it. Oxford University Press. Markus, Gregory B (1988). The impact of personal and national economic conditions on the presidential vote: A pooled cross-sectional analysis. American Journal of Political Science, 137–154. Maxfield, Sylvia (1998). Gatekeepers of growth: the international political economy of central banking in developing countries. Princeton University Press. McDowell, Daniel and David A Steinberg (2024). Black representation and the popular legitimacy of the federal reserve. European Journal of Political Economy 85, 102583. McNamara, Kathleen (1998). The Currency of Ideas: Monetary Politics in the European Union. Cornell University Press. Monnet, ´Eric (2024). Balance of Power: Central Banks and the Fate of Democracies. University of Chicago Press. Moschella, Manuela (2024). Unexpected revolutionaries: How central banks made and unmade economic orthodoxy. Cornell University Press. Mukherjee, Bumba and David Leblang (2007). Partisan politics, interest rates and the stock market: Evidence from american and british returns in the twentieth century. Economics & Politics 19 (2), 135–167. Onoda, Takuya (2024). ‘depoliticised’regulators as a source of politicisation: rationing drugs in england and france. Journal of European Public Policy 31 (5), 1346–1367. Panel, S , Emilien Paulis, J-B Pilet, S´ebastien Rojon, and Davide Vittori (2024). The lure of technocrats: A conjoint experiment on preferences for technocratic ministers in six european countries. Political Behavior 46 (3), 1961–1984. Pond, Amy (2021). Biased politicians and independent agencies. Journal of Theoretical Politics 33 (3), 279–299. Quinn, Dennis P and Robert Y Shapiro (1991). Economic growth strategies: The effects of ideological partisanship on interest rates and business taxation in the united states. American Journal of Political Science, 656–685. Rogoff, Kenneth (1985). The optimal degree of commitment to an intermediate monetary target. The quarterly journal of economics 100 (4), 1169–1189. Romelli, Davide (2022). The political economy of reforms in central bank design: evidence from a new dataset. Economic Policy 37 (112), 641–688. Scheve, Kenneth (2004). Public inflation aversion and the political economy of macroeconomic policymaking. International Organization 58 (1), 1–34. Selin, Jennifer L (2015). What makes an agency independent? American Journal of Political Science 59 (4), 971–987. Thatcher, Mark and Alec Stone Sweet (2002). Theory and practice of delegation to non-majoritarian institutions. West European Politics 25 (1), 1–22. Thatcher, Mark , Alec Stone Sweet, and Bernardo Rangoni (2023). Reversing delegation? politicization, de-delegation, and non-majoritarian institutions. Governance 36 (1), 5–22. Voelkel, Jan G , Michael N Stagnaro, James Y Chu, Sophia L Pink, Joseph S Mernyk, Chrystal Redekopp, Isaias Ghezae, Matthew Cashman, Dhaval Adjodah, Levi G Allen, et al. (2024). Megastudy testing 25 treatments to reduce antidemocratic attitudes and partisan animosity. Science 386 (6719), eadh4764. W¨alti, S´ebastien (2012). Trust no more? the impact of the crisis on citizens’ trust in central banks. Journal of International Money and Finance 31 (3), 593–605. Way, Christopher (2000). Central banks, partisan politics, and macroeconomic outcomes. Comparative Political Studies 33 (2), 196–224. Westwood, Sean J , Erik Peterson, and Yphtach Lelkes (2019). Are there still limits on partisan prejudice? Public Opinion Quarterly 83 (3), 584–597. ### Appendix #### A.1 Demographic Comparison with CCES Here we compare our survey with the Cooperative Congressional Election Survey (CCES). The CCES is a survey designed to be representitive of US adults. Table A1: Comparison of Means and Proportions Between CCES and Prolific Samples Variable CCES Prolific ```text Age (years) 50.39 46.1 College Degree 0.35 0.18 Political Party Identification 3.72 2.84 ``` Female (%) 53.2% 50.0% Male (%) 46.1% 49.1% ```text Other (%) 0.7% 0.9% ``` Household Income: $100,000 or more (%) 20.9% 11.5% Household Income: $50,000–$99,999 (%) 29.8% 33.3% Household Income: Less than $50,000 (%) 41.2% 53.9% ```text Household Income: Prefer not to say (%) 8.1% 1.3% ``` #### A.2 Full Model Specifications-Experimental Results ```text Individual Treatments Pooled CACE (Intercept) −0.370*** −0.372*** −0.370*** (0.073) (0.073) (0.074) ``` President Treatment 0.249*** (0.046) Democrats Win 0.215*** (0.046) Republicans Win 0.298*** (0.050) Pooled President Treatment 0.254*** (0.035) Recieved Treatment (CACE) 0.367*** (0.051) ```text CBI - Pre-treatment 0.659*** 0.659*** 0.667*** (0.024) (0.024) (0.023) Education 0.007 0.007 0.004 (0.009) (0.009) (0.009) Trust in Federal Reserve 0.067** 0.067** 0.064** (0.024) (0.024) (0.025) Trust in IRS −0.007 −0.007 −0.007 (0.023) (0.023) (0.023) Inflation Impact Perception 0.007 0.006 0.016 (0.025) (0.025) (0.025) State: Maine 0.824* 0.839* 0.825* (0.414) (0.425) (0.406) State: Minnesota 0.341 0.349 0.327 (0.227) (0.223) (0.217) State: Ohio 0.191+ 0.191+ 0.174+ (0.100) (0.099) (0.097) Loan Count (3 Loans) 0.174+ 0.175+ 0.176+ (0.091) (0.092) (0.091) Loan Count (5 Loans) 0.545 0.538 0.508 (0.486) (0.487) (0.517) Asset Count (7 Assets) 0.072 0.073 0.076 (0.061) (0.061) (0.063) Num. Observations 1535 1535 1535 R-Squared 0.497 0.496 0.494 ``` + p <0.1, * p <0.05, ** p <0.01, *** p <0.001 Table A2: Full Model Specifications of Figure 2 Table A3: Full Model Specification of Figure 3 CBI Support (Intercept) −0.090 (0.100) Co-Party Treatment −0.120* (0.052) Out-Party Treatment 0.116+ (0.060) CBI Pre-treatment 0.639*** (0.032) Education 0.005 (0.013) Trust in Federal Reserve 0.043+ (0.024) Inflation Impact Perception −0.026 (0.030) State: Colorado 0.346 (0.230) State: Kentucky 0.257 (0.210) State: Maine 0.895* (0.451) State: Michigan 0.233 (0.180) State: Minnesota 0.520+ (0.294) State: Virginia 0.217 (0.140) Vote in 2020: Biden 0.045 (0.049) Vote in 2024: Robert F. Kennedy Jr. 0.167 (0.126) Loan Count (3 Loans) 0.179 (0.111) Asset Count (3 Assets) 0.042 (0.067) Asset Count (7 Assets) 0.062 (0.083) Mortgage (2 Mortgages) 0.015 (0.049) Federal Reserve Knowledge (Low) 0.058 (0.049) Number of Observations 1008 R-Squared 0.450 + p <0.1, * p <0.05, ** p <0.01, *** p <0.001 #### A.3 Observational Analysis ```text Model 1 Model 2 Model 3 Model 4 Party Identification −0.022+ −0.022+ 0.002 −0.024+ (0.013) (0.013) (0.013) (0.013) Age −0.001 −0.001 −0.003+ −0.001 (0.002) (0.002) (0.002) (0.002) College Education 0.138* 0.144* 0.054 0.132* (0.058) (0.059) (0.056) (0.059) Income 0.009 0.010 0.004 0.014 (0.020) (0.020) (0.019) (0.020) Inflation Expectations −0.005+ −0.005+ −0.004+ −0.005+ (0.003) (0.003) (0.002) (0.003) Mortgage 0.028 0.031 0.042 0.033 (0.061) (0.061) (0.059) (0.062) Perceived Inflation Impact 0.168*** 0.168*** 0.083** 0.166*** (0.027) (0.027) (0.027) (0.028) Loan Count −0.008 −0.008 −0.007 −0.012 (0.044) (0.044) (0.043) (0.045) Asset Count 0.003 0.005 0.011 0.004 (0.014) (0.014) (0.013) (0.014) Makes Financial Decisions 0.310* 0.315* 0.290* 0.337* (0.127) (0.126) (0.113) (0.137) Grocery Shopping 0.141* 0.138* 0.107 0.132+ (0.069) (0.068) (0.066) (0.070) ``` Financial Literacy −0.002 (0.078) ```text Fed. Knowledge (index) −0.009 0.001 ``` (0.028) (0.026) Fed Knowledge (self assess) −0.030 (0.041) Trust in Government 0.043 (0.038) Fed Trust 0.237*** (0.034) ```text (Intercept) −0.804* −0.755* −1.112*** −0.862* (0.317) (0.326) (0.296) (0.371) Num.Obs. 1362 1362 1362 1326 R2 0.062 0.062 0.139 0.062 ``` + p <0.1, * p <0.05, ** p <0.01, *** p <0.001 #### CATE - Interest Rate Preference This table shows the full model specification for Figure 6 (1) Pooled Info. Treatment 0.220** (0.073) Interest Rate Preference −0.025 (0.038) CBI (pre Treatment) 0.658*** (0.024) Trust in Federal Reserve 0.070*** (0.019) Treatment × Interest Rate Preference 0.051 (0.046) Intercept −0.303*** (0.073) Num.Obs. 1471 R2 0.484 + p <0.1, * p <0.05, ** p <0.01, *** p <0.001 #### A.4 Pre-Post Election - Table A4: Pre- and Post-Election Sample Balance 0 1 Mean Std. Dev. Mean Std. Dev. Diff. in Means Std. Error ```text Party Identification 2.8 2.1 2.8 2.1 0.0 0.1 Age 46.1 15.9 48.7 15.3 2.6 0.6 Income 3.4 1.6 3.3 1.6 −0.1 0.1 Education 3.8 1.5 3.8 1.5 −0.1 0.1 ``` Table A5: Full Model Specification Pre-Post Election Party Identification −0.045 (0.066) Wave 2 0.281*** (0.050) Party Identification × Wave 2 −0.097*** (0.015) Num.Obs. 2258 R2 0.730 Respondent FE Yes + p <0.1, * p <0.05, ** p <0.01, *** p <0.001 #### A.5 Sankey Plots Figure A1: Sankey Plot - Full Sample: These plots show the movement of respondents pre- to post-treatment for the control group and the pooled treatment groups. All respondents are included in the analysis, N=1471 the for t 2 N=90 ost-treatmen p to analysis, pre- the ts in onden resp included of are t en emv mo Republicans the d w an sho crats plots Demo These Only arm. t artisans: P - treatmen ##### Plot h ey eac and ##### Sank A2: group trol Figure con ### Notes [^1]: “Tory MPs demand review of Bank of England Independence,” Bloomberg UK, May 19, 2024. [^2]: “Political ‘soap opera’ derails Australian central bank reform,” Financial Times, September 25, 2024. [^3]: In a tweet during his first presidency, Trump asked whether “the Fed chair was a bigger enemy to the US than Chinese President Xi Jinping.”“Donald Trump warns US Fed chair not to cut rates before the election,” Financial Times, July 17th, 2024. [^4]: Historically, Democratic Presidents were not shy in criticizing the Fed when pushing for a shift in monetary policy. The incident between President Johnson and Chairman Burns illustrates that disagreement is baked into the institutional independence of monetary policy (Bernanke, 2022; Kern and Seddon, 2024). [^5]: ‘Turkey’s Erdogan calls interest rates ’mother of all evil’,” Reuters, May 11th, 2018. [^6]: “Hungary’s Orban and his central banker,” Reuters, December 21st, 2011. [^7]: The literature on CBI conceives independence as a multi-dimensional concept. Generally, researchers differentiate between goal and instrument independence. While goal independence allows central bankers to choose their policy targets, instrument independence implies that monetary authorities can choose the instruments to achieve specific macroeconomic goals. Recent advances in measuring CBI have added important dimensions, including political interference in appointing and firing central bank leadership and financial autonomy of the central bank independence (Adrian et al., 2024; Cukierman, 1992; Romelli, 2022). [^8]: This advice, promoted by international financial institutions (Kern et al., 2019) and rewarded by international financial markets (Maxfield, 1998; Bodea and Hicks, 2018) has been widely adopted (Garriga, 2025b; Romelli, 2022). However, the degree of independence varies across countries, and many countries have reduced their central banks’ autonomy (Bodea and Garriga, 2023; Kern and Seddon, 2024). [^9]: The literature analyzing the role of central bank communication has primarily focused on households’ and firms’ inflation expectations (for a recent survey, see Blinder et al. (2024)). [^10]: For recent surveys, see Anson (2024) and D’Acunto et al. (2024). [^11]: Other research shows similar results regarding satisfaction with central banks, see Blanchflower and MacCoille (2009); Garriga (2025a). [^12]: Recent survey evidence reveals that citizens do not have a basic understanding of monetary policymaking [^13]: “Trump Touts Tariffs, Lashes at the Fed in Interview.” Bloomberg, October 15, 2024. [^14]: Although leaked documents on Trump’s plan to subordinate the Federal Reserve appeared in the Wall Street Journal, these plans were not prominently featured in public debates. “Trump Allies Draw Up Plans to Blunt Fed’s Independence.” The Wall Street Journal, April 24, 2024. [^15]: “Letter to Jerome Powell.” Elizabeth Warren, June 10th, 2024. [^16]: “The Importance of Central Bank Independence,” The White House, May 22nd, 2024. [^17]: Debates about the ongoing housing crisis in the U.S. have often linked its emergence to the Federal Reserve’s interest rate policies. Lawmakers from both parties have subsequently criticized the Fed’s rate decisions, blaming them for unaffordable mortgage rates and sparking discussions about the institution’s independence.“US housing crisis becomes a critical issue in the presidential election.” The Financial Times, August 17, 2024. [^18]: Anson (2024) documents that attitudes towards the Fed are very sensitive to different frames. [^19]: Note that this hypothesis slightly differs in syntax than the preregistered hypothesis. However, it aligns more closely with our pre-registered empirical specification. Further, we originally stated this hypothesis in two parts but condensed it here. [^20]: The survey was fielded on August 6th, 2024. This sample size should produce a minimum detectable effect of 10% of standard deviation at 80% power for Hypothesis 1. [^21]: Prolific also offers convenience samples. Our sample, in comparison, is based on quota sampling of respondents based on personal characteristics collected by Prolific before running the survey. In the Supplementary Appendix (Table ?? we show how our sample compares to the Cooperative Congressional Election Survey (CCES) sample. On average, respondents in our sample are slightly less educated, and report slightly less income. They also report being more democratic. [^22]: It reads as follows: “The Federal Reserve, established in 1913 to provide a safer financial system, has been both praised and criticized throughout its history. While supporters argue it has helped stabilize the economy during crises, critics contend it has contributed to economic inequality and market bubbles through its monetary policies. The Fed’s actions, particularly its handling of interest rates and quantitative easing, have sparked debates about its long-term impact on the U.S. economy.” [^23]: For those who received the final two treatments, we debriefed them with the following information: “Important: We may have made a claim earlier in the survey that stated either Trump or Harris were likely to win the election. While many polls are pointing in different directions, Political Scientists caution that it is very difficult to forecast an election outcome. This is especially true in August. The election outcome is still highly uncertain.” [^24]: We do not find this result surprising. See the divergent expectations for material interests and ideological biases discussed in the argument section. [^25]: The substantive effect is similar even when omitting the pretreatment CBI response from the right-hand side. [^26]: Note that we did not preregister this analysis in the left panel. [^27]: We asked: “Now, consider the role of interest rates (how expensive it is to borrow money on a credit card, for a car, or for a home AND how large a return you get on savings) in the economy. Which of the following situations is better for you, personally?” They could reply with ‘Interest rates go up (increasing the return on my savings), Interest rates go down (making credit card interest lower and loans cheaper), Interest rates stay the same, or I don’t know.” [^28]: In the Supplementary Appendix, we compare the basic demographics of those we did and did not recontact. We find no substantive differences in partisanship, age, gender, income, or education. [^29]: We distributed two surveys. One was for the control group that again did not provide any information regarding the counterfactual of reduced CBI. The other survey included the information treatment, excluding the electoral expectations text, for those who received one of the treatment arms in the first wave. [^30]: In subsequent analysis, we found that the election effect is not conditional on our own informational treatment. We suspect this is likely due to greater discussion of CBI in the press following the election. Among our recontacted respondents, 60% reported that Trump’s position on the Fed sounds, at least, familiar.