Argument and Contribution

At the national level, the United States struggled to effectively respond to the COVID-19 pandemic: federal policy was delayed and inconsistent, supply shortages were widespread, and political pressure undermined accurate public health guidance. At the state and local levels, however, there was a great deal of variation in terms of the capacity to respond to COVID. While indicators of state capacity often focus on “formal” indicators like institutional resources, staffing, and finances, translating formal capacities into effectively implemented policies is neither a simple nor an automatic process. 

In “Building the Plane While Flying,” Didi Kuo and Andrew S. Kelly draw our attention to the importance of informal indicators of public health capacity. These include strong relationships within and across government agencies, the embeddedness of health officers in local communities, and prior experience with responding to disasters, among other factors. The authors argue that local governments with strong informal capacity were better able to communicate with and learn from one another, as well as to gain the trust of community members, during the pandemic. Conversely, localities with otherwise strong formal capacities often failed to respond to the challenges at hand. This is because they were unable to effectively leverage their relationships and organizational networks.
 

Kuo and Kelly’s paper is informed by qualitative analysis of California’s public health institutions as well as in-depth interviews with health officers across seventeen Northern, Central, and Southern California counties. The interviews illuminate the concrete processes by which local governments responded (or struggled to respond) to the COVID-19 pandemic. One of the paper’s key contributions is to push us to conceptualize state capacity more broadly and to focus on the factors that drive not just policy development, but policy implementation. 

The Importance of Informal Capacity

The bulk of the paper disaggregates informal capacity into its various mechanisms and processes. Each of these proves to be crucial in explaining different pandemic outcomes at the local level. One such mechanism is coordination within local governments. To illustrate this, consider public health officers, who enjoy broad legal powers to protect public health as well as financial resources and personnel at their disposal. By law, officers possess significant capacities to mitigate health crises. Yet across the interviews, health officers reported that effectively implementing COVID-19 policy required their cooperation and communication with a host of actors, including the County Counsel (the county’s top lawyer), Chief Administrative Officer, and Board of Supervisors, which is charged with appointing health officers.
 

Closely related to intra-governmental coordination is the importance of autonomy, particularly in the face of political pressure. For example, boards of supervisors sometimes undermined the public guidance provided by health officers. (This guidance could range from the need to close schools to officers simply communicating truthfully with localities about COVID-19 risks.) Overcoming efforts by board members to ignore or muzzle officers required coordination between those actors who were more insulated from political pressure.

Another key component of informal capacity was prior experience responding to emergencies and California’s myriad of natural disasters, such as fires, floods, or mudslides. Health officers from more experienced counties noted their ability to draw upon established emergency procedures and partnerships. For example, some counties had previously cooperated with each other, as well as with independent agencies like the Red Cross, to provide aid and shelter to those affected by wildfires. These experiences — for which no amount of financial resources or personnel can substitute — served as templates to help coordinate COVID-19 policy responses.

Informal capacity also depended upon health departments effectively communicating with the public. Many departments initially lacked the infrastructure to do this, and therefore relied on cooperation with other actors like school superintendents, sheriffs, and community leaders. Some counties created toolkits to improve their community’s understanding of personal protective equipment (PPE) or even produced local TV shows. Still others scheduled conference calls with local hospitals, faith leaders, and nursing homes. Given that many of these communication efforts were improvised, public health officers stressed the importance of formalizing coordination between state and nonstate actors so as to improve emergency preparedness in the future.
 

In addition to coordination within local governments, effective policy-making and communication required coordination across governments. One such institution was the Association of Bay Area Health Officers (ABAHO), founded in the 1980s during the HIV/AIDS epidemic. ABAHO members had also coordinated policy responses to the H1N1 outbreak. These cross-county partnerships enabled early, rapid, and unified responses to the COVID-19 pandemic. By contrast, the authors find that regions without such networks faced greater challenges in developing and implementing public policy.

A final aspect of informal capacity is the social embeddedness of health officials in their communities. This includes partnerships with leaders of businesses and faith groups, teachers, and restaurant owners. Not only did these partnerships increase the scope of outreach, but they also often established relationships that had not existed beforehand. Gaining a foothold in local communities thus increased the likelihood that community members would support policies and enabled local governments to access hard-to-reach populations.
 

Kuo and Kelly’s analysis of informal capacity should give us pause when considering existing indices of public health preparedness; some of these have ranked the United States quite high despite its often ineffective responses to the COVID-19 pandemic. That this mismatch occurs is arguably a function of observers prioritizing formal capacities. For federal, decentralized countries like the US, informal capacities and relationships are essential not only for delivering services but for generating legitimacy and trust among those receiving services.

*Research-in-Brief prepared by Adam Fefer.

Motivation & Contribution

The Black Lives Matter (BLM) movement has sparked a domestic and global reckoning about racial injustice, especially with respect to police brutality. BLM mass protests — in particular, the misperception that these protests tend to be violent — have polarized much of American society, spawning pro-police countermovements like Blue Lives Matter. How can we understand the sources of public support for or opposition to racial justice movements like BLM? 

One suggestion is to look at the tactics used by protesters themselves, with the implication that more “disruptive” or “radical” tactics — for example, blocking highways as opposed to holding silent vigils — are key in determining how the public perceives BLM. In “How Police Behavior Shapes Perceptions of Protests,” Jasmine English, Ariel White, and Laurel Eckhouse provide evidence for a different answer: police tactics — in particular, whether protests were met with a heavy police presence — better explain why onlookers have perceived BLM protests as violent. After presenting data on the determinants of different police responses to protests, the authors conduct a survey experiment to explore how the presence of police at BLM protests affects public perceptions. 

The article deepens our understanding of the police as political actors who shape how Americans think about social movements. These perceptions are not solely determined by what protesters do or how the media portray them, but by how police respond. Worryingly for racial justice advocates, this means that violent police repression may fail to engender public sympathy for protesters; indeed, the mere presence of police leads onlookers to believe that protests are violent. Conversely, violent protest tactics are unnecessary for the public to perceive protesters as violent. The policy implications of these findings are that heavy police deployments at protests should be clearly justified and documented, given their potential to shape public opinion. 

Data & Findings

The authors assemble a dataset from over 1,000 BLM protests that occurred between 2014 and 2017. This includes details on protest tactics, timing (i.e., whether it was before or after dark), and size, as well as police tactics, including whether police were present, made arrests, or deployed crowd-control measures. The reader learns that BLM protests varied considerably in terms of size and place. So too did police responses vary, even in the face of similar protest messages and tactics. Police were present at nearly ⅔ of BLM protests, while at least one arrest occurred, and crowd control measures were deployed around one-fifth of the time. More generally, police responses ranged from using military gear, police joining in and kneeling with demonstrators, pushing elderly people to the ground, and undercover officers pointing guns at protesters.
 

Image

Protest Locations in the Contiguous United States
Note: Point sizes are scaled by approximate protest size. 

The authors run a regression analysis where the three dependent variables (DVs) or outcomes of interest are police presence, arrests made, and crowd control. The following independent variables (IVs) are found to be statistically significant in affecting all three DVs: highway blockages, protests occurring after dark, disruptive tactics (e.g., protesters changing themselves to objects), and large protests (i.e., over 1000 participants). However, the low predictive value of these IVs (also known as their R²) means the authors must find another way to understand how protests, police responses, and mass perceptions are linked in a causal way.
 

Image

Table 1: Protest Characteristics and Police Response

Survey Experiment

The authors then conduct a survey experiment to test the following two hypotheses: H1: protests with a heavy police presence are more likely to be perceived by respondents as violent. H2: protests with a heavy police presence will yield less support for the overall movement from respondents. Over 2700 people participated in the experiment, which entailed respondents receiving two different “treatments”: one group read a “vignette” about a peaceful BLM protest, while the other read about a peaceful protest with a large police deployment. In other words, the only difference is whether police were present.
 

Image

Figure 2: News Article Vignette

H1 is measured by asking respondents how much they agree or disagree — on a five-point “Likert” scale ranging from “strongly disagree” to “strongly agree” — with statements such as “the event was violent” or “protesters had violent intentions.” Similarly, H2 is measured by how much respondents agree with statements such as “I’d consider becoming involved in the protest,” “I support the protesters,” or “I’d post something positive about the protests on social media.” 

The authors find positive support for H1 but null support for H2. In other words, respondents who receive the treatment about police presence are more likely to view the protesters as violent, but this does not translate into reduced support for the movement. Their explanation for the null finding is as follows: “support” is a more medium- or long-term attitude, shaped by repeated “exposure” to protests over time. As such, a single exposure — in the form of reading a vignette — is unlikely to shift respondents’ support. By contrast, “believing something is violent” may be more easily shaped by a single exposure. Finally, the authors present exploratory (i.e., non-preregistered) data showing that the effects for white respondents in H1 are higher.
 

Image

Figure 5: Effects of Heavy Police Presence on Violence Perception

Image

Figure 6: Effects of Police Presence on Support for BLM Protesters

Ultimately, the article shows how police behavior, more than protester tactics, can shape public perceptions of social justice movements like BLM. These findings invite a reevaluation of important mass movements across American history — such as the Civil Rights Movement — where police responses played a pivotal role. A heavy police presence at otherwise peaceful protests can distort how the American public perceives constitutionally protected behavior, potentially undermining the accuracy that is necessary for democratic participation.

*Research-in-Brief prepared by Adam Fefer.

Motivation

Retaliation (or the threat thereof) is a central component of human behavior. It plays a key role in sustaining cooperation — such as in international organizations or free trade agreements — because those known to retaliate come to acquire a reputation of being hard to exploit. But how does the use and function of retaliation vary across cultures, and how does it interact with formal forms of punishment?

In “Cross-cultural differences in retaliation: Evidence from the soccer field,” Alain Schläpfer tackles these questions using data on retaliation from association football. Retaliation is simply defined in terms of fouling: player B retaliates against player A if and only if, after A fouls B, B then fouls A. Among other findings, Schläpfer shows that players from cultures emphasizing revenge are more likely to retaliate on the football field. This form of ‘informal punishment’ by players also interacts with ‘formal punishment’ by referees: retaliation by B is less likely when A is sanctioned with a yellow card. Schläpfer’s paper increases our knowledge of the causes and consequences of retaliation, while showing how informal cultural norms interact with the formal rules of football.  

Data

Schläpfer creates a data set of fouls committed over three football seasons (2016-2019) in nine of the world’s top professional men’s leagues. This includes the European leagues of Premier League (England), Serie A (Italy), Bundesliga (Germany), LaLiga (Spain), and Ligue 1 (France), as well as Série A (Brazil), Liga Profesional (Argentina), Liga MX (Mexico), and Major League Soccer (United States). The dataset comprises 9,531 games, 230,113 fouls committed by 10,928 unique perpetrators from 139 countries against 11,115 unique victims from 137 countries.

Because Schläpfer hypothesizes that being from more revenge-centric cultures explains on-field retaliation, the key independent variable is measured using a dataset from Stelios Michalopoulos and Melanie Meng Xue that identifies revenge motifs in a culture’s folklore. Examples of this include supernatural forces avenging human murders or animals avenging the death of their friends by humans. Schläpfer uses a host of other independent variables, such as country-level survey data about the desire to punish — as opposed to rehabilitate — criminals, which is also theoretically linked to revenge. As stated above, retaliation is measured in terms of fouls committed. Schläpfer shows that there is substantial variation in retaliation rates among players from different countries, from Gabon (8%) to Iceland (31%). Can the folklore in the country of origin explain the behavior of players on the field?
 

Image

Fig. 1. The share of fouls retaliated in soccer games (top) and the prevalence of revenge motifs in folklore (bottom). Both variables tend to have higher values for players and folklore from the Middle East, Central Africa, Eastern Europe, and parts of South America.
 

Findings

Retaliation:

Schläpfer finds evidence that players from cultures that value revenge are indeed more likely to retaliate for fouls. However, they are not more likely to commit fouls overall, cautioning us against conflating the concepts of retaliation and violence. Indeed, Schläpfer demonstrates that motifs of violence in a culture's folklore do not predict retaliation. Players are also found to be more retaliatory early on in a game, which is consistent with its use as a signal or aspect of one’s reputation. In other words, retaliation serves to deter future fouls. Victims of fouls also retaliate quickly. Indeed, retaliation rates are stable or slightly increasing during the first 30 minutes of a game, but then fall consistently thereafter.
 

Image

Evidence is also provided to show that retaliation deters future transgressions: perpetrators are less likely to foul again if victims retaliate for the initial foul. However, this deterrence finding is only observed when the perpetrator is from a revenge culture. In other words, for retaliation to support cooperation (the absence of fouls), players must share a similar cultural background.

Schläpfer’s findings hold even when soccer-related or socioeconomic factors are taken into account. Further, the paper considers, but finds little support for, alternative explanations of why retaliation varies. These include that some teams encourage players to retaliate more or employ more players from revenge cultures. Further, retaliation does not appear to be driven by emotions; otherwise, it would be less likely to occur after halftime when players have had a chance to cool down, but this is not the case.
 

Image

Informal and Formal Sanctioning:

Finally, Schläpfer analyzes the interaction between player retaliation and refereeing. Most importantly, retaliation is significantly less likely if a foul is sanctioned with a yellow card. This illustrates the theoretical principle of formal punishment “crowding out” informal punishment, such as religious excommunication, which carries greater weight than social shunning or police fines compared to peer pressure. Both retaliation and referee sanctioning are shown to reduce the frequency of repeated offenses by perpetrators, especially among players from revenge cultures. However, Schläpfer finds that formal sanctioning is roughly three times more effective than retaliation. This suggests that football referees are doing a better job managing conflict between players than players themselves. 

Schläpfer concludes by mentioning a few of the paper’s limitations. First of all, retaliation is measured only by what referees sanction. However, referees may miss crucial incidents for which retaliation is a response, such as Zinedine Zidane’s 2006 World Cup headbutt after a verbal insult (that was not sanctioned). This is important because individuals from revenge cultures are likely to be particularly offended by verbal insults. Second, the paper does not capture retaliation that occurs across games played by the same teams over time, particularly when rivalries and hostilities have intensified. Similarly, it does not account for preemptive retaliation that does not follow a foul. Ultimately, Schläpfer deepens our understanding of retaliation in a domain where many would expect it not to operate or to do so with minimal significance. The article impressively marshals large-scale data from both sports and cultural history to clarify the causes and consequences of retaliation.

*Research-in-Brief prepared by Adam Fefer.