Abstract
Patriotism is conceived of as a national-level concept. I posit that people hold similarly strong feelings toward their municipal area. Based on long-standing the- ories of patriotism in national politics, I show how local patriotism influences local politics. Using novel preregistered survey data from an online sample matched to nationally representative data in terms of gender, age, and race, I show that people have feelings of love, indifference, or hate toward their municipality. I also find a strong positive correlation between loving one’s municipality and partici-pation in local politics, civic participation, and trust in local government. I also conducted two preregistered survey experiments that show that priming feelings of love and/or hate towards one’s town strongly motivates the willingness to sac-rifice to solve local collective action problems. Specifically, stimuli that evoked these feelings made participants much more likely to donate the payment they earned from completing the survey to solve a town problem. These results show the crucial importance of local patriotism for understanding local politics.
Do citizens love their towns 1 the way they love their nations? If so, what impact does this local patriotism have on citizens’ attitudes and behaviors in local politics? Patriotism has almost exclusively been viewed as a national-level concept. 2 I posit that people hold similarly strong feelings toward their municipal area. Such feelings explain much about their participation in local politics, civic participation, trust in the local government, and willingness to sacrifice for the town. Based on long-standing theories of what motivates someone to being willing to incur the costs of collective action, I developed a theory of how local patriotism can influence behavior and attitudes, such as motivating participation in local politics. Empirically, I also show that local patriotism is a distinct form of attachment from previously measured concepts, such as national patriotism, and therefore needs to be studied.
I expect citizens with a greater love of their town to be willing to carry the cost of engaging in politics and civic activities. Conversely, if someone dislikes or even hates their town, they will have diminished feelings of civic duty. Such sentiments are necessary to explain any rational actor’s decision to participate in political activity that requires sacrifice (Riker & Peter, 1968). Hundreds of studies have examined attitudes and behaviors in local activities and civic participation (e.g., Titus, 1981), but none have conceptualized local patriotism as an explanatory factor. The persuasive empirical results below show that local patriotism is an important omitted variable in our understanding of local politics.
Since Aristotle, we have known that daily interactions in the agora (a local place of assembly, such as a marketplace) are crucial for understanding how people relate to their political world (see, e.g., Aristotle, 1988, vii, II, 2). Local interactions are the best chance for citizens to express themselves politically (DeSantis & Hill., 2004) because the nature of mass politics dilutes one person’s impact on national politics to the point of being ineffectual (Berry et al., 1993). However, depending on the town’s size (Oliver, 2000), an individual may affect political outcomes at the local level (Lassen & Serritzlew, 2011).
Nevertheless, participation in town politics and civic participation requires a willingness to get involved. The local level is where an impact may be more likely, but it is also the level where political participation is the least prevalent (Oliver et al., 2012). Low participation in town politics often leads to domination by a concerted minority mobilized by direct self-interest (see an example from housing policy in Einstein et al., 2018). As such, low participation levels may make outcomes unrepresentative of the general populace’s will in the town (Fung, 2006). Thus, understanding why people get involved in town politics is a crucial question for political science.
This topic is crucial to study because local government is a critical factor in determining the quality of life (Young & Kaczmarek, 2000). Still, many citizens ignore it by focusing on national politics or ignoring politics altogether. Local politics are well known to have a tremendous impact on citizens’ well-being (C´arcaba, Gonzalez, Ventura, and Arrondo 2017). Criminal justice, education, housing, and public health policies are mainly administered by the local government. A voluminous literature that an involved, committed citizenry will improve the local area and simultaneously improve many of the outcomes for the citizens themselves (see, e.g., Aldrich, 2012). A possible explanation for the lack of local political involvement may be a lack of willingness to sacrifice to achieve the town’s goals because residents do feel attached to the city (see Yang & Callahan, 2007). Generally, I posit that local patriotism should correlate with greater willingness to incur costs to help the town.
To investigate the impact of local patriotism, I preregistered two survey questionnaires, experimental protocols, and theoretical predictions. 3 Using novel survey 4 data matched to nationally representative data for gender, age, and race, I show that people have feelings of love, indifference, or hate toward their municipality. I also find a strong positive correlation between loving one’s municipality and participation in local politics, civic participation, and trust in local government. The survey results are augmented by the clear causality provided by two survey experiments with real cash consequences. The survey experiments show that priming feelings toward one’s town strongly motivates the willingness to make a financial sacrifice to solve local collective action problems.
In both experiments, participants were asked to identify the biggest problem facing their town. They were then asked to donate their $1 payment for taking the survey to help address that specific problem. The treatment groups were asked to think about their feelings of love and/or hate toward the town, while the control group was only asked to donate to help the town’s problem. The treatment group was much more likely to donate than the control groups in both experiments. This outcome shows the power of local patriotism in motivating behavior. These participants signed up to take the survey for money, yet thinking about their feelings toward their town led many more respondents to donate that money willingly. The survey experiments show that these results are not merely some form of social desirability bias or survey wording response bias.
This paper continues as follows: I examine how local patriotism has been concep-tualized in the literature, examine how national patriotism has affected political attitudes and behaviors, and state several theories about how local patriotism can be influential. I then describe my preregistered predictions; detail the experimental procedures, data, and methodology; describe the results, and conclude with a discussion of the findings.
Conceptualizing Local Patriotism
While no prior research in political science considers the impact of local patriotism on politi-cal participation, civic participation, trust in local government, and collective action, similar concepts have been investigated. There is literature on local patriotism in the field of history examines local uprisings in the past—particularly the 1700s and 1800s—whereby townsfolk opposed the nation-state, showing that they care more about their locality than their na-tion (e.g., Ashton, 2010). More directly, a few articles have examined how Europeans view their region compared to their nationality in human geography. For example, researchers have investigated how Spain’s Basque region citizens are more likely to be proud of being Basque than being Spanish (Melich, 1986). Also, Gnatiuk (2017) examined local patriotism in Ukraine and compared it with national patriotism among Ukrainians. However, none of these studies examined the impact of local patriotism on political attitudes or behaviors.
In sum, while some research has examined elements of local patriotism, the predic-tions that arise from this research has not been tested directly. As such, I now examine the research on national patriotism and how it affects attitudes and behaviors, and I use this information to delineate the possible effects of local patriotism on attitudes and behaviors at the local level.
Patriotism at the National Level
The extensive literature on how national-level patriotism affects national-level politics serves as a helpful starting point to understand how local patriotism may affect local politics. The literature on patriotism almost universally assumes that this concept concerns the love of the nation-state. Crucial to this research, the literature has consistently found that na-tional patriotism correlates with participation in national-level politics, such as voter turnout (Straughn & Andriot, 2011). This finding leads to the expectation that local pa-triotism will also correlate with participation in local politics.
While early research on patriotism often conflated it with nationalism, modern research uses a more nuanced approach. It defines patriotism as a positive feeling of appreciation for the nation and a willingness to sacrifice for its greater good (Davidov, 2009). In contrast, nationalism is described as a feeling that one’s nation is superior to others. 5 Further, the literature presents nationalism as a harmful component of democratic citizenship (Schatz et al., 1999). Crucially, while nationalism is correlated with bigotry and xeno-phobia, patriotism is not (De Figueiredo & Elkins, 2003). Patriotism instead correlates with greater participation in politics and a willingness to make sacrifices for the nation’s good, for example, by joining the military or even not cheating on taxes (Konrad & Qari., 2012). Nationalism, however, does not correlate with these beneficial forms of volunteerism (Richey, 2011).
Patriotism encompasses multiple categories, which are most prominently represented by constructive patriotism and constitutional patriotism. For example, multiple factors of patriotism were uncovered in Kosterman and Feshbach (1989) by conducting factor analyses of survey answers to questions about feelings toward the nation. Researchers posit that there is a constructive patriotism that involves individuals who see problems in their society and work toward solutions, which often require sacrifice in some way for the greater good (Schatz et al., 1999). Constitutional patriotism features a deep feeling of at-tachment and allegiance to the nation’s raison d’ˆetre and essential documents, but it does not devolve into blood and soil nationalism (Muller, 2007). The critical point for constitu-tional patriotism is that anyone can be a full member of society if they have an allegiance to its foundational national aspirations, essential documents, and civic culture. By doing so, constitutional patriotism welcomes immigrants and historically marginalized subgroups as patriotic members of society.
Local patriotism is related to constructive patriotism in that its focus is on helping the town. It is not related to nationalistic or blind patriotism because it is measured below with a question that asks if the respondent loves, likes, neither/nor dislikes, or hates their town. The question used below does not ask about opinions of the town’s superiority or general relationship with other towns because the underlying concept of local patriotism is not conceptualized in this research as having anything to do with superiority. The underlying concept of local patriotism is also related to constitutional patriotism because anyone who moves to the town could develop a love for it. The uniqueness of this measure of local patriotism is that it is used in this research for the first time to measure the impact of these concepts at the local level on critical political phenomena.
These streams of research conclude that patriotism is not to be confused with na-tionalistic or xenophobic belligerence toward outgroups and minorities, and it features a willingness to sacrifice to solve the nation’s problems. These forms of patriotism motivate political participation, civic participation, and trust without necessarily devolving into na-tionalism (Gangl et al., 2016). Extending these ideas to the local level, local patriotism should not cause hostility toward surrounding towns or subgroups in the town but should engender a willingness to bear the costs required to participate in and engage with one’s town for its betterment. Using this understanding of how patriotism interacts with politics at the national level, I now present several ideas about how local patriotism can shape citizens’ participation with local politics.
Local Patriotism and Citizens’ Attitudes and Behaviors
The first research question is: What influence does local patriotism have on local political participation? The early research into political participation demonstrated that civic duty is crucial to understanding the cost–benefit analysis for civic participation and political partici-pation. The fundamental logic explicated in Downs (1957) should hold for local participation and participation in national politics. Simply put, one person typically does not make much of a difference, and the cost of involvement almost always outweighs the plausible possibility of achieving a political goal. Whether attending a meeting, donating to a candidate, or emailing a City Council member, one additional action from an individual is not likely to determine a change in public policy or electoral results.
However, the cost–benefit analysis for citizens who love their town and want to improve it is different than those how do not love the town. For those who love their town, sacrifice includes the instrumental benefit and the psychological benefit of helping improve the town they love (Riker & Peter, 1968). These benefits likely outweigh the cost of participa-tion for many individuals who express local patriotism. Thus, knowing the foundation for the psychological desire to help the town is crucial for understanding political participation. I posit that local patriotism is a profound motivator that alters the cost–benefit analysis of participation, with greater patriotism equating to greater participation. Thus, I present prediction 1:
Prediction 1. Local Patriotism Correlates with Greater Participation in Local Pol-Itics
The second research question is: What influence does local patriotism have on civic participation? Patriotism may correlate with civic participation for an intuitive and straight-forward reason: The more people care about their town, the more they will volunteer in ac-tivities that promote its improvement. At their core, almost all civic participation activities hinge on improving the local situation. Citizens who care more about the local community should engage with it more often. Furthermore, those who dislike or hate their town would be less likely to sacrifice their time and effort to engage in voluntary and other civic activities to benefit that town. Thus, prediction 2 is the following:
Prediction 2. Local Patriotism Correlates with Greater Civic Participation
The third research question asks what influence local patriotism has on trust in local government. Decades of studies have shown that the relationship between political trust and attitudes toward government is complex. It is difficult to discern whether people like the government and therefore trust it or trust the government and subsequently like it. Corre-lational survey analysis does not go far in untangling the intertwined causal relationships of trust and affection. While noting the lack of causation, I expect that local patriotism should significantly correlate with political trust because it is doubtful that people would like the city’s government that they could not trust.
Furthermore, while the survey evidence provided below cannot identify causation di-rectly, it is essential to know that this relationship exists. Many public health campaigns, for example, demonstrate the unique role that trust plays in getting members of the community to cooperate with public health officials. If we can find a way to boost local patriotism, we may find an avenue to increase trust in local political institutions and compliance in these campaigns that require cooperation with the town. Based on these ideas, prediction 3 is as follows:
Prediction 3. Local Patriotism Correlates with Greater Trust in Local Government
The fourth research question centers on the influence that local patriotism has on solv-ing collective action problems. These problems are well-known social traps that feature a large amount of free-riding from individuals unwilling to sacrifice for the greater good (Olson, 1965). Local patriotism potentially offers a solution to social traps that are created by collective action problems. The simple idea is that if people love their town, they will be more willing to trust that others will not cheat in the iterated prisoner’s dilemmas that give rise to social traps (Rothstein, 2005). Further, a greater love of town may inspire people to be more likely to sacrifice short-term self-interest for the town’s benefit. With greater local patriotism, we should see a process similar to what is often referred to as a patriotic act by those who sacrificed for their nation. Thus, prediction 4 is the following:
Prediction 4. People Primed with Local Patriotism are More Likely to Contribute to Solving Local Problems
Experimental Procedures
To examine whether local patriotism affects the willingness to sacrifice for the town, I con-ducted two survey experiments. Experiment 1 was conducted on the same sample from which the survey data were derived in October 2020. Experiment 2 was from a different online sample, with slightly different stimuli explained below in April 2021. In both exper-iments, participants were first asked to write what they considered is the biggest problem facing their town. For example, in Experiment 1, approximately 15% indicated that traf-fic was the biggest problem in their town, while another 12% identified crime as the biggest problem. The central idea of both experiments was to test whether inducing local patriotism influences the desire to sacrifice for the town by donating money to help solve its biggest problem.
Both experimental designs are based on the emotion induction methodology in psy-chology (see Lench et al., 2011). This methodology is used to create stronger emotions in a treatment group than in a control group (Siedlecka & Denson, 2019). Experimental research in psychology shows that after having participants think about emo-tional attachments, their subsequent actions will be more likely to be influenced by them (see Baumeister et al., 2007 for a critical review). The difference be-tween the treatment and control group in the current study derives only from the randomly assigned inducement of the emotional attachment to their town. Having the treatment group think about their feelings towards the town prioritized that emotional attachment in deciding whether to donate while the control group was not necessarily thinking about these feelings. Importantly, if participants do not have feelings of emotional attachment, then stimuli to prompt such feelings will not influence their behavior. Based on its long history in psychology, vetted by dozens of psychometric experiments (see a meta-analysis in Westermann et al., 1996), this design and method are appropriate for this research question and theory.
In Experiment 1, the treatment group was primed to consider their feelings of love and hate toward the town. If, after considering their feelings about the town, this group is more likely to donate to help solve town problems, a causal relationship would be revealed between local patriotism and the willingness to sacrifice to solve collective action problems. In this case, the sacrifice is the participant’s money earned from filling out the survey. Note that while $1 may not seem like a substantial sacrifice, it was sufficient to motivate the respondent in the Prolific sample to fill out the survey and indicative of an extensive enough reward to motivate effort. “You are making $1 for taking this survey. “You are making $1 for taking this survey. Would you like to donate this $1 to help your town solve the problem that you just listed above? If so, please click Yes, Donate $1 below. If not, please click No, Do Not Donate $1.”
In Experiment 2, a third of the sample was randomly assigned to be asked to think about their feelings of love toward the town. Another third of the sample was randomly as-signed to be asked to think about their feelings of hate toward the town. The control group was asked for the same donation, but without the stimulus to prime their feelings of love or hate toward the town. All three groups were then asked to donate the money ($1) that they had earned by taking the survey. The two treatments and control stimuli are respectively listed below: “You are making $1 for taking this survey. “You are making $1 for taking this survey. “You are making $1 for taking this survey. Would you like to donate this $1 to help your town solve the problem that you just listed above? If so, please click Yes, Donate $1 below. If not, please click No, Do Not Donate $1.”
Randomization produced balance across a broad set of pretest variables (see Appendix Tables A1 and A8. 6 ) The experiments used deception, which was revealed after the survey was completed. The respondents were told that the money was not donated to the town and that they could donate it themselves. 7
Data and Methods
The data analysis was conducted using preregistered survey questionnaires with survey ex-periments. The first sample included 500 US citizens who Prolific recruited. The survey was conducted online, and respondents were matched to nationally representative age, gen-der, and race Census data. The respondents were paid $1 for completion. 8 A total of 524 respondents started the survey, but only 468 finished it. Incomplete surveys were excluded from the analysis. See Appendix Table A2 for a table of summary statistics for all variables. I included one manipulation check to ensure that the participant read the survey (Kane & Barabas., 2019). All respondents responded correctly to the instruction, “Please select 3 from the responses below.”
The second survey used in Experiment 2 was sampled about 6 months later than the first using the same online methodology and survey company, Prolific. However, this was an online convenience sample that was not matched to nationally representative benchmark data. The point of the second survey was to replicate Experiment 1 at a different time with a different sample.
Dependent Variables
I followed past researchers in measuring political participation on a scale of possible activities. The logic of using such a scale is that there are many ways to participate in politics. Simply measuring voter turnout, for example, is not sufficient to measure the diversity of ways in which the public can be involved in politics. Further, because people can participate with larger or smaller amounts of time, it is best to weight participation by active involvement. Local Political Participation was measured with an additive scale of expressed activity in the following political activities: never (0), once or twice (1), or more than twice (2). Survey participants were asked to mark how often they have engaged in the following local political activities: sent an email or letter to a local politician, commented on social media about local politics, discussed local politics with friends or family, volunteered for a local election campaign, donated money to a local campaign, and attended a meeting/event on a local political issue. Each question specified that it was about local politics, not to confuse the respondent that this might be a about a national-level political issue or person. For example, the question on sending an email or letter to a local politician, was worded as “Sent an email or letter to a local politician”.
As with political participation, there are also many ways to be involved civically. As such, I also use a scale to measure civic participation weighted by the level of participation. Civic participation was adapted from an additive scale of expressed civic activity suggested by the National Research Council (2014). It is measured with agreement with the statements that the respondents have not done this (0), have done it a little (1), or have done it a lot (2). The participants were asked, “Have you ever spent time participating in any activity below in the last 12 months?” The list of activities included involvement with a church or religious group; parent-teacher association; neighborhood organization/homeowners associ-ation; youth sports; Boy Scouts/Girls Scouts or similar youth groups; Rotary, Masons, Elks or Moose Lodge, Lions, Kiwanis, or similar group; Chamber of Commerce or local business group; cleaning-up litter project or similar beautification group; and other community group not listed.
I adapted a question for local political trust from the American National Election Studies (ANES) study that has been frequently used to measure national-level political trust (ANES, 2016). Local Political Trust was measured with the question, “Now speaking of the municipality that you listed above, how often do you trust the local government to do the right thing?” Answers were coded as always (4), most of the time (3), about half the time (2), sometimes (1), and never (0).
Collective Action Problem Solving was measured with a question about participants’ willingness to donate the fee they earned for completing the survey to help the town solve an important problem. Responses were coded as 1 if a participant chose to donate and 0 if they did not.
Local Patriotism Variable
The 2012 ANES included a question on patriotism at the national level, which I adapted for the local level by asking, “Now speaking of the municipality that you listed before, how do you feel about the town you live in?” Responses were coded as follows: hate it (−2), dislike it (−1), neither like nor dislike it (0), like it (1), or love it (2). Interestingly, not many demographic factors correlated with local patriotism, although it seemed to increase with greater age and income. (See Appendix Table A3 for regression models of correlates of local patriotism.)
I have no way to verify the content validity of the wording of this local patriotism question, which centers on a novel concept. While it is not easy to assess content validity, the question does have face validity, in that it is plausible that respondents can reasonably assess their feelings toward their town. And although it is measured with a single question and not a scale of questions, it is intended to measure local patriotism generally. Single-item measures also have some psychometric advantages. For example, having a single item reduces the common method variance problem, a type of systematic error shared among similarly measured variables. 9 Due to the shared common understanding of the terms like, dislike, love, and hate, the question should be able to determine the level of local patriotism for this sample and be able to test my hypotheses.
Figure 1 shows a skew in the distribution toward more positive feelings about the town. Indeed, the modal category indicates that around half of the sample liked their town. How-ever, it is essential to note that a full 25% did not express any positive feelings toward their town, and 3% hated their town. This outcome confirms that local patriotism exists; real feelings of love and hate are distributed across a nationally representative sample toward the town they live in. As such, it brings into focus the need to study such feelings and their implications for participation in town politics, civic participation, trust in local government, and a willingness to sacrifice for collective action problems. . This bar graph shows the distribution of local patriotism.
This is important to show because, while we may not expect every person to have similar ideas of love or hate toward their town, the distribution of this unmeasured variable is currently unmeasured and, thus, unknown. Intuitively, because each town is unique and can have broad variations in the quality of life and provisions of resources, there will be wide variations in attitudes toward it. Importantly, negative experiences with a town can also range widely, encompassing challenges such as corruption and conflicts with police.
The Tiebout model suggests that people can move if they do not like their town, and having such a choice works as a form of pressure on towns to provide a better quality of life (Tiebout, 1956). However, newer research suggests that people are closely tied to their communities and do not move as freely as the Tiebout model suggests, even if they do not like it. Consequently, some residents may love their town while others hate it, and a large group exists between these extremes.
I also expect that there are underlying psychological propensities to express patriotism and that these tendencies will be positively correlated between the expression of patriotism at the national level and the local level. Beyond psychological propensities, various factors have likely socialized the respondent to have stronger or weaker patriotic feelings, and the distinction between national and local governments will not necessarily erase the effects of socialization.
However, while several reasons support the expectation that national and local pa-triotism will be positively correlated, citizens can distinguish between local and national governments. Their patriotism may therefore differ across these distinct levels. Since there is only one federal government but 37,000 municipalities in the United States, I expect a much broader variation in attitudes toward towns than the national government.
Figure 2 shows a scatterplot of national patriotism measured on the same 5-point scale from hate to love of America as the local patriotism variable. Figure 2 is jittered so that the dots are viewable, and it shows that a correlation exists between local and national patriotism. Pairwise correlations show that these two variables correlate at .340 with a p-value of less than 0.001. While a significant positive correlation does exist, plenty of people in the sample felt differently about the nation and their town. This difference is apparent from the number of people who, for example, love America but hate their town. This scatterplot shows the relationship of local patriotism and national pa-triotism.
In the regression models below, I controlled for national patriotism to address the un-derlying psychological propensities and political socialization processes that influence general opinion toward the government. This allows me to isolate the unique effect of local patrio-tism on the dependent variables.
Control Variables
I control for national patriotism to separate the general psychological propensity and the political socialization of a generalized tendency to like government. The models below show local patriotism’s separate and unique impact by including national patriotism in their re-gression models. The question for national patriotism was taken from the 2012 ANES and was worded as “Now speaking of the United States of America, how do you feel about it?” Responses were coded as follows: hate it (−2), dislike it (−1), neither like nor dislike it (0), like it (1), or love it (2).
Similarly, I wanted to control for general tendencies to trust government and other people to accurately measure the impact of local patriotism on trust in the local government. For interpersonal trust, I used the ANES question, “Generally speaking, how often can you trust other people?” Responses were coded as follows: always (4), most of the time (3), about half the time (2), once in a while (1), and never (0). The question about trusting the national government was also from the ANES: “Now speaking of the United States of America, how often do you trust the federal government in Washington DC to do the right thing?” Responses were coded using the same scale as for interpersonal trust.
For the political participation and civic participation models, I wanted to control for a general desire to participate in politics and care about political actions. First, I measured voting in the 2020 presidential elections by asking, “Whom do you plan to vote for in the 2020 Presidential elections?” I coded responses as 1 for those who answered they did not plan to vote and 0 for those who said they planned to vote. Additionally, I control for interest in politics with the question, “How closely are you following politics?” Responses were coded as follows: very closely (3), somewhat closely (2), not too closely (1), or not at all (0).
I also controlled for Education, Age, Income, and whether the respondent was Black, Hispanic, or Female because socioeconomic status variables have been shown to have a massive impact on individuals’ political participation, civic participation, and trust in gov-ernment (Verba et al., 1995). Additionally, I controlled for how long the respondent resided in the town because longer experiences with a town could plausibly lead to greater trust and attachment (Kang & Kwak, 2003). The variable Residency Tenure was measured in years.
Methods
Local Political Participation and Civic participation were count variables that indicated the number of activities a respondent was involved in. To conduct the statistical analysis with these dependent variables, I used a Poisson regression model. Trust in Local Government was measured with an ordinal variable, and I used an ordered logistic regression model for its analysis. The decision to donate or not to the town in the experiments was a dichoto-mous variable, and I used a logistic regression model to measure the effect of assignment to treatment. For all these models, the modeling choice was robust to alternative specifications, such as pairwise correlations and ordinary least squares regression models (e.g., see Appendix Tables A4 and A5 for pairwise correlations and ordinary least squares regression models of these dependent variables and local patriotism showing similar results).
I use data from the first survey for the correlational analysis testing predictions 1, 2, 3, 4, and 5 below because of the validity derived from it being matched to national repre-sentative data. For this reason, I omit correlational analysis from the second convenience sample below. However, Appendix Table A7 shows that the same correlations predicted by predictions 3, 4, and 5 obtain in this convenience sample; although the p-value for Local Participation is above .05, it is signed in the expected positive direction. Additionally, in survey 2, Local Patriotism showed a similar distribution as Figure 1 (see Appendix Figure A1) and has a pairwise correlation with National Patriotism at 0.3243 with a p-value < 0.001.
Results
Prediction 1
Determinants of Participation in Local Politics.
Note: Cells represent unstandardized coefficients and standard errors for Poisson regression models for determinants of participation in local politics.
+p<.10, *p<.05, **p<.01, ***p<.001.
The control variables match expectations from prior research over what is influencing political participation. Having more interest in politics, the length of residency in the town, greater education and income, and being female raised the propensity to participate in poli-tics, while not voting in national elections correlates with less participation in local politics. The other control variables were not statistically significant.
Prediction 4
Determinants of Civic Participation.
Note: Cells represent unstandardized coefficients and standard errors for Poisson regression models for determinants of civic participation.
+p<.10, *p<.05, **p<.01, ***p<.001.
The control variables match expectations from prior research over what is influencing civic participation. Having more interest in politics, the length of residency in the town, greater education, and being Black correlates with civic participation, while age correlates with lower likelihood of participating in local politics. The other control variables were not statistically significant.
Prediction 3
Determinants of Trust in Local Government.
Note: Cells represent unstandardized coefficients and standard errors for ordered logistic regression models for determinants of trust in local government.
+p < .10, *p < .05, **p < .01, ***p < .001.
The control variables do not have much impact on trust in local government. Hav-ing more interpersonal and national level political trust correlates with more trust in local government, while age correlates with lower trust. The other control variables were not sta-tistically significant.
Prediction 4
Prediction 4 was that local patriotism would correlate with a greater willingness to solve local collective action problems. Specifically, this prediction was measured by the respondent being willing to donate their payment for completing the survey ($1) to solve a problem that they identified about their town. This was tested with two survey experiments. This box plot shows the mean and 95% confidence intervals of the willing-ness to donate to help solve the town’s biggest problem for the treatment and control groups.
In Experiment 1, shown in Figure 3, the treatment group, which was prompted to consider their feelings of love and hate toward the town, 18% said they would donate the money to help their town. In the control group, which received no prompts about feelings, only 3% said they would donate the $1 to the town. The six-fold difference between the treatment and control groups is significant at a 0.001 level. 11
This result shows that stimulating feelings of local patriotism by having the respon-dents think of their feeling towards the town generated a willingness to incur costs to solve collective problems. When thinking about their feelings towards their town, one in six of the treatment group respondents were willing to incur an actual monetary cost to help the town. Without being spurred to think about their feelings towards their town, only one in 30 participants in the control group were willing to sacrifice for the town. This outcome demonstrates that local patriotism is a powerful motivator for solving collective action prob-
Experiment 2, Figure 4 shows that no one in the control group was willing to pay the $1 they received for the survey to help their town’s problem. In the treatment group that stimulated love of the town, we see that 8% were willing to donate to the town. The difference between these two conditions is statistically significant with a p-value < 0.001. This confirms prediction 6 and replicates the finding in Experiment 1. This box plot shows the mean and 95% confidence intervals of the willing-ness to donate to help solve the town’s biggest problem for the treatment and control groups.
Interestingly, the stimuli which provoked hate of town also saw 5% of its participants willing to donate the $1 to help the town. This is also statistically significantly different from the control group with a p-value < 0.001. Of note is that the difference between the treatment conditions stimulating feelings of hate of town and love in town is not statistically significant. This suggests that stimulating any feelings towards the town will generate more willingness to sacrifice to resolve collective action problems. Perhaps engendering emotions 12 To consider how the treatment effects vary by how people felt about their town, I divided the sample into groups that either liked (4) or loved their town (4), or those who had no opinion (3), disliked (2) or hated their town (1). There were 128 participants in the Less Local Patriotism Group and 340 participants in the More Local Patriotism Group. The results of box plots shown in Appendix Figures A2 and A3 show that the treatment effect is similar for both groups. Interestingly, no one in the Less Local Patriotism Group donated to help the town’s problem in the control condition, but around 14% donated when stimulated with local patriotism. This suggests that even people who have a basic dislike of the town can be motivated through patriotic appeals.
Toward the town in and of itself is motivating sacrifice when compared to a control group in which no emotions were stimulated.
These results clearly show that if you can prime local patriotism, it may lead to more willingness to sacrifice for the town. There are several ways to prime patriotism, but perhaps the chief way is incentivizing a willingness to work with the town. Richey (2007) finds that experiments done in Japanese towns show that incentivizing working with the town could increase trust in the town. This finding suggests that a similar incentive process can engen-der local patriotism by facilitating working with the town.
Conclusion
I find that local patriotism exists; people do have feelings of love or hate for their town. The distributions of national patriotism and local patriotism are correlated, but not strongly. Thus, an underlying psychological predisposition or political socialization process that pre-dicts someone expressing a positive attitude toward government does not entirely explain the phenomenon of local patriotism. One or more additional factors distinguish people’s feelings toward local government from their feelings toward the national government. Even after controlling for national patriotism, I found that local patriotism matters greatly in local politics. A robust and clear correlation was apparent between local patriotism and political participation in local politics, civic participation, and local government trust. The correlations were robust to multiple model specifications and data analysis. Although causal-ity cannot be obtained for predictions 3, 4, and 5 from cross-sectional survey data, these pre-registered strong results show that these relationships need much more study. Crucially, I show that local patriotism is a distinct concept from previously measures of attachment, such as national patriotism, and therefore needs to be studied.
I also conducted two survey experiments that showed that priming feelings toward
One’s town are a powerful motivator of sacrifice to solve local collective action problems. In the first experiment, stimuli that elicited feelings of love or hate toward the town encouraged a six-fold increase in the number of participants willing to donate their payment to solve a town problem. Exposure to treatment increased donation rates from 3% to 18%. The second experiment found that the stimulating feelings of love of town led 8% of participants to donate their fee while stimulating feelings of hate led to 5%. These conditions were statistically significantly different from a control group that did not stimulate feelings toward the town. These results show that more work in political science is needed to understand how feelings toward the local area affect local politics.
In the political science literature on participation in local politics and civic participa-tion, local patriotism has been mostly ignored, but these results clearly show that it should not be. These findings help elucidate the role of civic education in promoting positive images of the town. The common concerns about patriotic education becoming nationalistic, xeno-phobic and jingoistic are not applicable for local patriotism (see MacMullen, 2012) because there is no local military or typically harsh xenophobic relationships with nearby towns. The clear findings of this research, including the survey experiments, show that one of the best ways to mobilize political action at the local level is to engender feelings of local patriotism.
Future research can examine how these results are moderated and mediated by local context, such as town size, It is true that this is not expected finding. One interesting finding was that prompting hate of the town elicited more donations. I expected that priming hate of the town would have the opposite reaction, or potentially have no reaction in terms of promoting donating to help the most prominent problem in the town. Perhaps getting the participant to think about their dislike of the town, makes them also think about prominent problems in the town, and thereby elicits more willingness to donate money to help these problems, which are now top of mind. This is a possibility that future researchers should investigate. Future research would also benefit by analyzing these potential subdivisions of local patriotism that are similar to those found in national patriotism. The literature has made several distinctions between the multiple forms of patriotism. This includes de-tailed literature on constructive patriotism, blind patriotism, nationalistic patriotism, and constitutional patriotism. These might also exist at the local level.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Notes
Appendix 13
Test of Balance between Treatment and Control Groups. Note: Cells represent means, and difference between the means with a significance test for difference in means between the treatment and control group. +p<.10, *p<.05, **p<.01, ***p<.001.
(Treatment)
(Control)
(Difference)
Local patriotism
0.711
0.888
0.178**
National trust
2.286
2.262
−0.025
Civic participation
2.122
2.031
−0.092
Local trust
3.000
3.004
0.004
Nonvoter
0.094
0.052
−0.042*
Interpersonal trust
3.183
3.283
0.100
Local participation
4.364
4.748
0.384
National participation
5.461
6.031
0.570
National patriotism
0.813
0.798
−0.014
Residency tenure
18.443
19.373
0.931
Hispanic
1.047
1.073
0.026
Black
0.132
0.142
0.010
Education
3.766
3.657
−0.109
Income
4.643
4.601
−0.042
Age
4.464
4.635
0.171
Female
1.545
1.506
−0.038
Observations
235
233
470
Summary Statistics.
Variable
Mean
Std. Dev
Min
Max
N
Donate
0.109
0.312
0
1
468
Treatment
0.498
0.501
0
1
468
Local patriotism
0.799
0.915
−2
2
468
Local participation
4.554
3.292
0
14
457
Civic participation
2.076
2.921
0
18
458
Local trust
3.002
1.026
1
5
467
National patriotism
0.806
1.112
−2
2
468
Interest
3.167
0.762
1
4
468
Nonvoter
0.073
0.26
0
1
468
National trust
2.274
0.932
1
5
467
Interpersonal trust
3.233
0.817
1
5
468
Residency tenure
18.906
16.101
0
98
468
Hispanic
1.06
0.237
1
2
468
Black
0.137
0.344
0
1
468
Education
3.712
1.085
1
6
468
Income
4.622
3.031
1
9
468
Age
4.549
1.618
1
7
468
Female
1.526
0.5
1
2
468
Determinants of Local Patriotism. Note: Cells represent unstandardized coefficients and standard errors for ordered logistic regression models for determinants of local patriotism. +p<.10, *p<.05, **p<.01, ***p<.001.
Variable
1
S.E.
2
Ideology
0.011
0.068
Nonvoter
−0.660
0.367
Trump approval
0.047
0.147
Republican party ID
0.017
0.305
Residency tenure
0.008
0.006
0.007
0.006
Hispanic
−0.517
0.358
−0.552
0.360
Black
−0.359
0.258
−0.274
0.263
Female
0.049
0.177
0.046
0.179
Age
0.150**
0.058
0.140*
0.060
Income
0.081*
0.033
0.074*
0.033
Education
0.104
0.092
0.087
0.093
Cutpoint 1
−3.092***
0.731
−3.217***
0.874
Cutpoint 2
−1.140
0.658
−1.256
0.814
Cutpoint 3
0.047
0.653
−0.058
0.810
Cutpoint 4
2.556***
0.666
2.462**
0.819
N
468
468
chi2
31.183***
34.761***
Correlation Matrix. Note: Coefficients and p-values in parentheses for pairwise correlations.
Local Patriotism
1.00
National patriotism
0.30 (0.00)
1.00
Local participation
0.12 (0.01)
−0.06 (0.20)
1.00
Civic participation
0.13 (0.01)
0.21 (0.00)
0.39 (0.00)
1.00
Local trust
0.41 (0.00)
0.26 (0.00)
0.08 (0.09)
0.18 (0.00)
1.00
Donate
0.16 (0.00)
0.05 (0.30)
0.10 (0.04)
0.14 (0.00)
0.09 1.00 (0.05)
Determinants of Local Participation, Civic participation and Trust in Local Government. Note: Cells represent unstandardized standard errors for ordinary least squares regression models for determinants of local participation, civic participation and trust in local govern-ment. +p<.10, *p<.05, **p<.01, ***p<.001.
Variable
Pol. Part
S.E.
C.E.
S.E.
Trust
S.E.
Local patriotism
0.556**
0.175
0.285+
0.152
0.416***
0.049
National patriotism
−0.276
0.144
0.513***
0.126
0.144***
0.040
Intercept
5.980***
0.957
−1.643
0.837
1.688***
0.268
N chi2
457
458
467
Willingness to Donate Survey Participation Fee to Fix Town Problem. Note: Cells represent unstandardized coefficients and standard errors for ordered logistic regression models for determinants of willingness to donate survey participation fee to fix town problem. +p<.10, *p<.05, **p<.01, ***p<.001.
Variable
1
S.E.
2
S.E.
Treatment
1.860***
0.397
1.807***
0.401
Local patriotism
0.676**
0.219
Nonvoter
−0.316
0.777
Intercept
−3.346***
0.360
−3.976***
0.447
N chi2
468 29.668***
468 41.252***
Correlations in Survey 2 with Local Patriotism. Note: Cells represent unstandardized coefficients and standard errors for Poisson and ordered logistic regression models. +p<.10, *p<.05, **p<.01, ***p<.001.
Variable
Participation
S.E.
Civic participation
S.E.
Trust
S.E.
Localpatriotism
0.009
0.022
0.176**
0.057
0.622***
0.141
Intercept
2.348***
0.147
−0.629
0.394
Cut point 1
1.482
0.945
Cut point 2
3.749***
0.958
Cut point 3
5.116***
0.983
Cut point 4
8.266***
1.094
N
259
259
260
chi2
0.173
9.852**
20.123***
Test of Balance between Treatment and Control Groups in Experiment Note: Cells represent unstandardized coefficients and standard errors for a multinomial lo-gistic regression model for determinants of assignment to either the love treatment, hate treatment or the control group. +p<.10, *p<.05, **p<.01, ***p<.001
Variable
1
S.E.
Hate treatment heightLocal patriotism
−0.066
0.157
Civic participation
0.098
0.053
Local trust
−0.121
0.140
Nonvoter
−0.273
0.385
Interpersonal trust
0.214
0.176
Local particpation
−0.104*
0.049
National patriotism
−0.023
0.144
Residency tenure
−0.005
0.010
Hispanic
0.216
0.401
Black
−0.827
0.508
Education
−0.274
0.158
Income
−0.074
0.052
Age
−0.007
0.107
Female
−0.082
0.261
Ideology
−0.050
0.078
Intercept
2.956
1.641
Love treatment heightLocal patriotism
−0.065
0.162
Civic participation
0.058
0.053
Local trust
0.061
0.143
Nonvoter
−0.289
0.393
Interpersonal trust
0.087
0.176
Local participation
−0.058
0.046
National patriotism
−0.106
0.143
Residency tenure
−0.018
0.010
Hispanic
−0.187
0.417
Black
−0.295
0.479
Education
−0.172
0.158
Income
−0.024
0.051
Age
0.066
0.105
Female
−0.346
0.261
Ideology
0.014
0.079
Intercept
2.763
1.632
N
256
chi2
25.220
This bar graph shows the distribution of Local patriotism from the Survey 2 convenience sample.
This box plot shows the mean and 95% confidence intervals of the willingness to donate to help solve the town’s biggest problem for the treatment and control groups for those who neither like or dislike, dislike or hate their town.
This box plot shows the mean and 95% confidence intervals of the willingness to donate to help solve the town’s biggest problem for the treatment and control groups for those who like or love their town.