Introduction

Citizens often deal with senseless bureaucratic requirements in their interactions with the state, experiencing frustration and inconvenience in the form of administrative costs. They manifest in a variety of ways, such as lengthy wait times, confusing instructions about public services, repetitive requests of information, or stress and discrimination for engaging in stigmatized programs (Herd and Moynihan, 2018). Conceptualized as administrative burdens, these difficulties are a significant challenge for citizen-state interactions as they can impede access to important benefits and services, even in self-service digitalized public services (Madsen et al., 2022), like the request for the “auxílio emergencial” (Schymura, 2020).

In recent years, there has been a growing interest in understanding how governments can reduce administrative burden. Besides the purportedly favorable universal adoption of digitized services, these new forms of interaction allow for studies about citizens’ relationships with public authorities (Madsen et al., 2022), as they have a new role for changing perspectives of burdensome requirements. One approach focuses on the communication process between citizens and public leaders.

Public leaders can change experiences by mediating and buffering costly relationships between citizens and bureaucracies, creating different perspectives, guiding them through learning processes about which rules (and which do not) work, and making them understand and deal with difficult claiming processes (Behn, 1998; Bellé, 2013). For example, Sobral et al. (2020) found that how leaders frame their messages influences citizens’ evaluation of government and their emotional responses to the situation at hand. However, in contexts of crisis like the pandemic, some types of leaders can be requested and emerge more prominently than others (McLellan, 2022). While citizens may expect effective information from administrative leaders, they may also want the reassurance from political leaders that crisis will be resolved (Mau et al., 2022; Wang et al., 2021).

Existing research has largely ignored the potential of administrative leadership to alter perceptions of onerous experiences (Moynihan, et al., 2012), creating a gap in knowledge at the intersection of two different literatures, namely public leadership (Mau et al., 2022) and administrative burden (Herd and Moynihan, 2018). Besides that, public leadership scholarship has predominantly focused on political leaders’ response to crisis (Crayne and Medeiros, 2021), leaving other forms of leadership overlooked (Ospina, 2016; Couto et al., 2022, Mau et al., 2022). Even within public leadership literature, the intersection of different types of leaders with citizens remains scarce. For example, the effect of different communication standards related to the constant ministerial changes during Bolsonaro’s term (Lima, 2022) or to the resignment of several public health directors in the Trump’s administration during the COVID-19 outbreak (Mau et al., 2022) was not extensively examined from the perspective of citizens.

Given the uniqueness of crises, leaders’ responses to them, and the extreme and varying experiences among citizens (Lagowska et al, 2020; Peci, Avellaneda and Suzuki, 2021), we explore how different types of leadership, employing the same message, impact citizens’ psychological experience of burdens, namely their levels of stress and stigma when requesting the “auxílio emergencial”. The association of leadership with administrative burden has not been extensively explored (Moynihan et al., 2012), especially considering the diversified contexts of leadership studies (Backhaus and Vogel, 2022). Therefore, we chose the psychological burden as our focus of analysis given that leaders frequently use sensemaking to target citizens' negative emotions (Hattke et al., 2020), creating a logical link between leaders' communication and citizens' emotional responses to bureaucratic interactions.

Moreover, we have reasons to believe that the extent public leaders influence their followers’ ability to comprehend (Maphinda, 2017) and alter the perception of administrative burden depends on citizens’ ideological lens (Sobral et al., 2020). It is already well documented that citizens’ evaluation of government’s performance is biased in favor of the leaders of the same political orientation (Crayne and Medeiros, 2021; Lau et al., 2022). Bass (2008) exemplifies the phenomena among right-wing Israeli voters who evaluate right-wing leaders more friendly and task-oriented than left-wing leaders. However, the experience of negative interactions for requiring the emergency aid (Schymura, 2020; Marins et al., 2021) has not been observed yet when the same right-wing oriented citizens are the beneficiaries themselves. Evidently, in a context of high polarization, ideological affinity can drive the perception of leader effectiveness (Lau et al., 2022; Jost, 2017). However, given some citizens’ judgements about targets of welfare (Goenka and Thomas, 2022), we wonder if the burden they experience might also be changed by their ideological lens.

To test the above rationale, we have conducted a survey experiment to compare how those leaders impact stress and stigma experienced by beneficiaries of the “auxílio emergencial”. Our findings partially support our hypotheses in a sense that right-wing leaders’ communication diminished the stress of rightists and left-wing citizens experienced higher stress and stigma when faced a message from all types of leaders.

By conducting this research, we aim to offer some important contributions to the literature. First, to extend prior knowledge about ways of mitigating administrative burden through public leaders’ communication. Second, to examine whether political affinity with the leader (Jost, 2017) moderates citizens’ response to burdensome interactions (Tummers et al., 2015). And finally, to add specific nuances to the understanding of the relationship between citizens and distinct public leaders in a new context, which is the digital application for a public benefit during the pandemic.

Administrative Burden

Administrative burdens are costs that emerge from citizens’ interaction with the state which have a potential to diminish citizen’s take-up rates. It can be distributive, in a sense that it affects some groups more than others; consequential, as it can limit and restrict access to public services and benefits; and constructed, by being the product of political choice embedded in public policies (Moynihan et al., 2015; Peeters, 2020; Herd and Moynihan, 2018). Although conceptually very close, administrative burden is distinct from organizational rules and red tape as it takes citizen’s experiences as onerous (Burden et al., 2012).

Authors regard administrative burden as a three-dimensional concept, with compliance, learning, psychological facets (Herd and Moynihan, 2018). Compliance burden can be measured by the time a citizen spends in a queue or by the documentation demands. Learning burden is present when one needs to find out more about applying for a public benefit, and psychological burden can be characterized as an increase on citizens’ stigma, stress, frustration, and/or loss of autonomy for applying for an unpopular program (Halling and Baekgaard, 2022).

With the advent of digitalized interactions, administrative burden scholarship has turned its attention to new settings of citizen-state interactions (Lindgren et al., 2019; Peeters and Widlak, 2018; Peeters, 2023) and their relationship with authorities (Madsen et al., 2022). For instance, authors claim that those who live in economic scarcity may experience higher stress because digitalized interactions lack personal assistance, have less space for empathy, clash with standardized operations with individual needs, and may even entail unfair treatments (Madsen et al., 2022; Peeters, 2023). Ultimately, those interactions bring high possibilities of increasing psychological costs for potential beneficiaries, and that is where public leadership comes in.

Public Leadership and Communication Messages

Effective communication can serve as a valuable tool for public authorities to mitigate administrative burdens (Madsen et al., 2022). For example, an experiment that provided students with information packets that included a list of schools given their grades, tuition costs and a voucher for free college applications, made low-income students 46 percent more likely to attend a selective institution than a control group that did not receive any information (Herd and Moynihan, 2018). Onerous interactions are not only an opportunity for providing valuable information, but also impose a responsibility on public leaders to provide guidance and relief.

However, the messenger is just as important as the communication message. Our take is that different profiles of public leaders conveying the same message might play a different role on citizens. While most scholars recognize the impact of public leadership on burdens, there is still debate about which types of leaders result in which outcomes and the extent of this impact (Backhaus and Vogel, 2022). For example, Moynihan et al.’s (2012) seminal work shows that leaders’ inspirational motivation and communication can alter the perception of costs by facilitating an open learning environment and reducing follower’s perception about senseless rules. The authors observed that agency leaders, appointed heads at the city level who have some professional experience, could enhance goal clarity and political support, and indirectly reduce red tape.

Few would disagree about the importance of administrative leaders and their ability to influence followers. Administrative leaders are staff with formal authority in public agencies (Ospina, 2016), uphold professional competence, are more concerned with policy implementation, tend to assume a position of “neutrality” and rely more on bureaucratic control mechanisms (Moynihan et al., 2012). Their position belongs to the administrative subfield of the core administration (federal, state, or local governments, public agencies) (Backhaus and Vogel, 2022) and they contrast to the elected leader, to whom we call “political”, in a sense that they are primarily focused on achieving organizational objectives and implementing policies (Van Wart, 2013).

Administrative leaders and the implementation of policies for which they are responsible for are not neutral (Marins et al., 2021), but their political ideology may not be as salient for followers as elected representatives. Indeed, political leaders are legitimized by elections, are more transparent in their political leanings, and may feel they have more demands from followers (Lees-Marshment and Jones, 2018). For example, evidence suggests that while conservative citizens tend to disfavor welfare benefits (Goenka and Thomas, 2022), right-wing politicians are also more tolerant of barriers (Aaroe et al., 2021).

Based on our analysis, we posit that administrative leaders are more inclined to adhere to established protocols and procedures within their organization, whereas political leaders tend to craft their messages based on public opinion and preference, thereby creating a distinct association with their persona. The effects of these two types of leadership on citizens can be variable, thus leading to our initial hypothesis:

H1: Political leaders have a stronger impact on citizens’ psychological burdens compared to administrative leaders.

Political Ideology

In general terms, political ideology is a shared set of beliefs about the order of society and how it can be achieved (Jost, 2017). It provides the lens through which citizens interpret their experiences and the environment they are inserted. The literatures of leadership and political ideology usually focus on the different types of messages that each leadership profile signals. For instance, citizens’ ideological differences require that political representatives vary their communication patterns to construct messages in an ideologically congruent fashion with their target citizens (Jost and Sterling, 2021). However, we wonder what the effects on citizens are when different leaders convey the same message. For instance, Backhaus and Vogel (2022) discovered that laissez-faire leadership, a hands-off and “neutral” style of leadership for delegating authority, has shown to be ineffective compared to any type of leadership, suggesting that “being neutral” is not what is expected from leaders.

The choice of neutrality over the polarization between right and left-wing was based on a careful analysis of the political and administrative landscape. We acknowledge the historical and current predominance of right and left political spectrums in Brazil. However, we chose to explore neutrality as a variable of interest because the prominent political leader during the time of the emergency aid concession (2019-2020) leaned towards the right, not the left. Consequently, any mention of the political leadership that won elections in 2018 would be associated with the right-wing orientation. Hence, our second hypothesis check if this is the case:

H2: Right-wing leaders have a stronger impact on citizens’ psychological burdens compared to neutral leaders.

We also know that citizens’ pre-existing political orientation has a clear impact on how they respond to a leader's communication during a critical event, such as the COVID crisis (Sobral et al., 2020). According to Sobral et al. (2020), leftists showed higher trust in government when political leaders framed more specific messages and showed pessimistic behavior regarding the pandemic, but the same did not happen with rightists. At the same time, when citizens’ ideological lens match with the leaders’, there tends to be a positive and emotional response from the audience. In the case of right-wing citizens, conservative leaders may have a diminishing effect on their psychological burden, as they may perceive them as more politically aligned with their values and interests. Our hypotheses aim to verify this:

H3: Citizens’ political ideology moderates the relationship between leaders and psychological burdens, in a sense that right-wing citizens would feel less burdened when they receive a message from right-wing leaders.

Recent events have shown that neutral leaders became a counteract to denialist right-wing leaders who refuse to follow scientific guidelines, causing confusion and even harm to citizens. Unfortunately, we have also seen a context of contamination of leaders, whose neutrality and expertise have been questioned (Lissardy, 2020), leading to distrust and confusion among the population. Therefore, our fourth hypothesis tests if the ideological neutral position still holds:

H4: Ideologically neutral leaders have the same impact on psychological burden for both right and left-wing citizens.

Methodology

Context and study design

To help small business owners, households, informal and unemployed workers, the government has granted them an amount of R$ 600 (six hundred reais) during the economic shut down provoked by the pandemic. However, several issues emerged since public leaders announced and started to grant the benefit. Problems with the mobile app of the public bank (“Caixa Tem”), inconclusive personal data, lack of proper communication, fraud, denied requests and other multilevel problems (Cardoso, 2020; Marins et al., 2021).

Given that the financial aid was granted based on previous administrative data, we looked at the psychological dimension of administrative burden, i.e, stress and stigma. These items were used in an exploratory analysis by Thomsen, Baekgaard and Jensen (2020) to measure the psychological costs of citizens’ coproduction. To give an idea, discussion groups about the emergency aid on social media grew immensely between 2019 and 2020, for exchanging information and sharing their experiences to receive the aid. There were many requests for help and stress over the process of receiving the aid. Moreover, at the time of the survey, we noticed that those negative psychological factors were very salient among beneficiaries who were still applying for the benefit and many of them were afraid of scammers and personal data leakage on social networks.

Methodologically, we designed four treatment conditions:

These treatment conditions were then combined in two dummy variables to gain more statistical power to run the comparisons, as follows:

(i) Leader type: political leader (i) with the right-wing political leader (ii) (coded as Leader type= 0); and administrative leader (iii) with the right-wing administrative leader (iv) (coded as Leader type = 1).

(ii) Ideology of the leader: neutral political leader (i) with the neutral administrative leader (iii), coded as Ideology of the leader = 0; and right-wing political leader (ii) with right-wing administrative leader (iv), coded as Ideology of the leader= 1.

We chose to describe the political and administrative leaders as right-wing (treats (ii) and (iv)) to reflect reality, given that the government in power at the time of the survey (2020) identified itself within this political spectrum. We have shown to participants a speech of the leader with some personal description and manipulated only the description of the type of leader, while keeping the statement unchanged across all treatment conditions. The original descriptions were in Portuguese and can be provided by request.

The leaders’ speech repeated across all the treatment conditions and participants were randomly assigned to one of them. We designed the discourse based on real-world speeches by the former Brazilian President, Jair Bolsonaro (political leader), and the president of Caixa at that time, Pedro Guimarães (administrative leader) available on YouTube. Although we did not mention their names in the manipulation, the discourse was made to show a standardized but targeted speech. The leader’s statement is described below:

"The current topic is the emergency aid. We have already paid around 33 million people. We had an expected number of people, but it went beyond that. And there are a few million who are complaining. The criteria for access to the Emergency Aid benefit are provided by law. The bank can only release the funds after personal data validation by the Administration. As soon as the company that has the data completes the analysis of the registration and sends the information to the bank, we provide an update on the situation on the website/app and release the funds. The application process remains the same. It undergoes improvements from time to time, but all within the expected timeline. The Emergency Aid is R$ 600, and only a few people are left to be paid. Regarding the individuals who fell into requirements, who are under analysis: some are genuinely entitled, some made mistakes, and others are not entitled, but that's okay. So, we are helping not only because of the pandemic, but it is also a government issue."

Variables’ measures

Therefore, we had two different treatment variables:

(a) Leader type, a dummy variable, in which 0 = political leader and the right-wing political leader groups, with 66 observations (49% of our sample), and 1 = administrative and the right-wing administrative leader groups, with 69 observations (51% of the sample);

(b) Ideology of the leader, a dummy variable in which 0 = “neutral” leaders (political and administrative) with N=74 (55% of the sample), and 1 = the right-wing leaders groups, N= 61 (45% of the sample).

We measured the impact of each scenario on people’s stress (a sample item reads “I am concerned that the process of receiving the emergency aid will affect my physical and psychological health negatively”) and stigma (a sample item reads “Requesting the emergency aid can project a bad image of myself to others”). They were measured using four items each (Cronbach’s alpha: ) in a 5-points Likert scale, where 1 is low and 5 is high levels of stress/stigma. The full description of the variables and all their measurement items are described on the following table.

To isolate the effects of leadership on our dependent variables, we control for several individual factors, such as gender, region of housing, educational level, marital status, and working condition (unemployed, individual entrepreneur, informal worker, and person whose family monthly income per person does not exceed half a minimum wage). In the demographics, we also asked people’s political orientation, which went from left-wing to right-wing orientation in a five-points scale. To gain more statistical power in the regression analysis, we combined the center-right ideological position with the right one, and the center-left position with the left one, keeping only three orientations (center = 0, right wing = 1, and left = 2).

Sampling strategy

The profile of beneficiaries of the emergency aid in Brazil is quite diverse. Initially, the program was aimed at informal workers, individual microentrepreneurs (MEI), freelancers, the unemployed, and low-income individuals. With the extension of the program and the opening of new registrations, other categories of workers, such as ride-hailing drivers and artisanal fishermen, were also included as potential beneficiaries. In addition, there is a significant portion of beneficiaries who are single-parent family heads, women who provide for their families, and people with disabilities. Given the application for the aid happened online through a mobile app during the pandemic, we identified and monitored some locus of organization of beneficiaries on social media. For example, Samuels and Zucco (2013) identified Facebook as a vast source for experimental subject recruitment. We systematically applied the survey on groups created by citizens who asked for the financial aid on Facebook.

Several groups were opened since the beginning of government announcement of the financial aid and, together, they have more than 250 thousand members and intense engagement and posts every day. On Facebook, we mapped and collected data on 29 groups which description was about “auxílio emergencial”. The group with more members had 193 thousand accounts and the smaller one had almost 2 thousand people. As a quantitative study, we tried to ensure the representativeness of the sample based on the beneficiaries’ profile, but the recruitment costs were prohibitive, so we used Facebook’s targeted recruitment via advertisement in the webpage of our management school. Given this process, our sample should be characterized as a non-probabilistic convenience sample.

The sampling took place from June 9th to August 25^th of 2020. Although we opted for an online convenience sample, we had an initial target of 250 responses based on the number of the four survey conditions and the statistical power it offers. In the end, we collected 296 answers. After excluding participants who did not complete the survey, we ended up with 135 participants (45% of response rate). Participants were informed they would be asked questions about leadership and checked an informed consent that ensured confidentiality of their identity. We used Qualtrics software to collect the data.

Descriptive Results

Our total sample was composed by 98 women (73%). 52 people (39%) declared they had a left-wing political orientation while 31 people (23%) had a right-wing orientation. 79 respondents were single (59%), 56 have completed the high school (42%) and 83 were from the southeast region (62%). We used chi2 tests to analyze the balance of our independent variables and the two non-balanced distributions were gender and region (p<0.05). We did not consider the gender imbalance a major problem for the analysis, as we know that women are the main target group of cash transfer programs, however, the higher representation of the Southeast region may indeed limit the generality of our findings.

Table 3 presents the pairwise correlations between the dependent and independent variables included in our regression analysis. We can observe some significant associations between stigma with our treatment variable of ideology of the leader (neutral x right-wing leaders) (r = 0.17, p<0.05), with citizens’ political orientation (r = 0.18, p<0.05), education (r = 0.32, p<0.05), and work status (r = 0.19, p<0.05).

The stress average in the political leader condition was 3.59 (SD= 1.08), while the average for the administrative leader was higher, with 3.66 (SD= 0.97). The averages for stigma were the opposite, with the political leader group showing higher mean (M= 2.16, SD= 0.97) and the administrative leader group with a lower one (M= 2.14, SD= 1.09). We conducted an ANOVA to examine the effect of the type of the leader (Political x Administrative) on stress level, but the results indicated that there was no statistically significant difference between the groups (F(1, 133) = 0.15, p>0.05) for Stress and for Stigma (F(1, 133) = 0.02, p>0.05).

Regarding the groups of right-wing versus neutral leaders, the stress average of the 74 participants exposed to the neutral leaders was higher (M= 3.75, SD= 0.93) than the 61 individuals in the right-wing leader conditions (M= 3.48, SD= 1.11). The ANOVA results do not show a significant difference in stress between the both groups (F(1, 133) = 2.42, p>0.05). However, the means of stigma for the ones in the right-wing leader treatment (M= 2.34, SD= 1.14) was higher than the neutral leader treatment (M= 1.99, SD= 0.90) and the ANOVA results showed a marginally significant difference between them (F(1, 133) = 3.92, p= 0.056).

Normality Assumption

In order to verify the normality of the dependent variables, the Jarque-Bera test was conducted, which is a goodness-of-fit test that determines whether the sample data do or do not have skewness and kurtosis that correspond to a normal distribution. In this case, both for the stress variable (skewness = -0.61; kurtosis = -0.18) and the stigma variable (skewness = 1.08; kurtosis = 0.65), the null hypothesis of the test statistic of the Jarque-Bera test was rejected, indicating that the sample data do not have a normal distribution (JBstress = 8.61, p<0.5; JBstigma = 28.9, p<0.01).

To achieve approximately normal distributions, monotonic transformations were carried out on the variables. The stress variable was squared, while the logarithmic function (log10) was applied to the stigma variable. These transformations are frequently used in the statistical literature to normalize data that are not normally distributed. As described by Hair et al. (2010), "transformations such as these are common in statistical analysis and can be applied to meet the assumptions of normality when necessary" (p. 123). Thus, the success of the transformations was verified with a new Jarque-Bera test, which confirmed the normality of the dependent variables (JBstress = 5.67, p>0.5; JBstigma = 3.53, p>0.05). Next, a normal P-P probability plot serves as a diagnostic tool, where a normal distribution of residuals aligns closely with the diagonal line of the plot.

Homoscedasticity

The homoscedasticity assumption concerns the distribution of residuals. For the data to be homoscedastic, the residuals must be uniformly distributed across the range of predicted values, resembling a "shotgun blast" of points. This uniform distribution ensures that the variance of the errors is constant. This was tested by observing the residuals scatterplot, which showed no obvious pattern, with points equally distributed above and below zero on the X-axis and to the left and right of zero on the Y-axis.

Additionally, a Breusch-Pagan test was performed as part of the statistical analysis. This test is widely used to check for the homogeneity of variances of residuals in regression models. In the context of this study, the Breusch-Pagan test was applied after multiple regression modeling to check if the variance of the errors was constant in relation to the independent variables. The test results did not indicate the presence of heteroscedasticity in the residuals for either the stress variable (p = 0.70) or the stigma variable (p = 0.63), suggesting that the assumption of homogeneity of variances was met.

Multicolinearity test

The assessment of the multicollinearity assumption in linear regression analysis was conducted using the Variance Inflation Factor (VIF) test. This test is crucial for identifying the presence of multicollinearity among the independent variables, which can affect the precision of the model's coefficient estimates. In the present study, all variables were subjected to the VIF test, and the results showed values close to 1 for each of them. This finding suggests that there is no significant evidence of multicollinearity among the independent variables, indicating that the regression coefficient estimates are reliable and not distorted due to the presence of this phenomenon. This analysis reinforces the robustness of the linear regression model used in this study to explain the relationship between the dependent and independent variables.

Regression Analysis

Political x administrative leaders

We conducted some linear regressions to measure the direct effect of the political x administrative leaders’ manipulation on the levels of stress and stigma. As observed on models 2 and 3 of Table 5, there was no significant effect of the type of leader on stress (β= 0.001 and β= -0.003, p> 0.05). For stigma, results are similar. The OLS predictors of stigma point to non-significant effects of the independent variables, as described on models 5 ( = -0.15, _p>0.05) and 6 ( = -0.16, _p>0.05) (H1).

As an additional analysis, we have graphically compared the stress and stigma averages per citizens’ ideological orientation for political and administrative leaders, as observed on Figures 3 and 4. The graph on the left shows the differences of stress between the leaders for center, right, and left oriented citizens. We can see that the confidence intervals overlap among most of political orientations, which confirms that there is almost no evidence of a relationship between the type of leader (political x administrative) and levels of stress.

However, from Figure 3, citizens with a left-wing orientation showed significantly higher stress for the administrative leader condition ( = 0.30, p<0.05) in comparison to the political leader scenario.

The graph on the right of Figure 3 shows that the confidence intervals for the three political orientations also overlap, indicating there is no statistically significant difference of stigma between the groups. Overall, although the interaction terms indicate that the direction of the effects is as expected - i.e., right-leaning citizens tend to experience less stress ( = -0.28, p>0.05, model 3) and stigma ( = -0.85, p>0.05, model 6) when led by a right-wing leader and left-leaning citizens tend to experience more - the lack of significant effects for leadership type precludes us from drawing firm conclusions in this regard.

Right-wing x Neutral leaders

When we make the ideological orientation of the leader salient, we noticed a diminishing impact on citizens’ level of stress. The regression results returned a significant effect of the right-wing leaders (type_leader = 1) on stress compared to the neutral leader (type_leader = 0), as observed on model 2 (β= -0.45, p< 0.05) of Table 5, confirming our hypothesis H2. However, when we add the interaction term to the model (model 3), this relationship remains negative, but loses the significance. It is possible that the interaction term is capturing some of the variation that was previously being explained by the main effect, leading to a decrease in its significance.

Theoretically, it is possible for leaders' communication to alleviate citizens' stress burden when their ideology is congruent with that of the leaders, but figure 4 suggests that all citizens (center, right and left-wing) had their stress levels reduced when the leader showed a conservative orientation. For the neutral leader, having a more liberal orientation (left) significantly increased citizens’ level of stress ( = 0.31, p<0.05). Still from Figure 4, the comparison also showed a significant positive difference of stigma for left-wing citizens when the leader is from a right-wing spectrum ( = 0.59, p<0.05), but a diminishing level of stigma for conservative citizens ( = -0.48, p<0.05), as theoretically expected and supporting the H3.

To investigate our hypothesis 4, we compared the levels of stress and stigma for the neutral leaders for each ideological position, as observed on Figure 4. The interaction of neutral leaders and right-wing citizens did not show any effect on the stress level ( = -0.02, p>0.05), but the interaction between neutral leader and left-wing citizen did ( = 0.31, p<0.05). The level of stigma did not significantly alter for any interaction of neutral leader and citizens ideological orientation. So, we could not confirm our hypothesis 4.

Moreover, education and the type of work (low per capta) can be good predictors for the levels of psychological burdens, with significant coefficients in all the models of Table 5. These results can be associated to the idea that more educated voters are less likely to feel stressed in public-citizen interactions but may feel more stigmatized for claiming a public benefit.

Note 1: Standard errors in parentheses.

Note 2: * p < 0.05, ** p < 0.01, *** p < 0.001

Note 3: Leader type and Ideology of the leader are dummy variables for the treatment conditions. Leader type: 0 = Political leader; 1= Technical leader Ideology of the leader: 0 = Neutral leader; 1 = Political and technical Right-wing leaders

Note 4: Political orientation of the citizen: 0= Center; 1= Right-wing; 2= Left-wing

Final Remarks

The aim of this research was to understand the impact of different types of public leaders on citizens’ administrative burden when requesting financial aid during the COVID-19 pandemic. Overall, we observed some insightful findings. Although we could not find any significant difference between political and administrative leaders, right-wing leaders diminished citizens’ level of stress to request the financial emergency aid when compared to “neutral” leaders, confirming that, for leaders, the salience of ideology can play an important role to diminish negative experiences.

To have a deeper understanding of which groups of citizens benefited from those leaders’ communication, we looked at their political ideology. We identified that more conservative citizens experienced a reduced level of stigma when right-wing leaders communicate with them but, when citizens reported a left-wing orientation, they experienced more burdens for all types of leaders. It is reasonable to consider that leftists feel more stressed and stigmatized given the conservative orientation of former representatives. Also, these results suggest that political ideology not only alters citizens’ perception of administrative burden but also impacts the way citizens respond to different types of leaders.

Our study could advance in some important fronts. Theoretically, we could bring the debate about public leadership into the administrative burdens’ literature and advance on the importance of administrative leaders for digitalized citizen-state interactions. Specifically, we argue that public leaders may have an important role on buffering citizens’ perception of costs to deal with the government, but this impact is conditioned on citizens’ ideological affinity. Although we could not find robust differences between political and administrative leadership, we can attribute it to the lack legitimacy of the former leaders to entitle beneficiaries of auxílio emergencial.

The idea of the antileader, according to Couto et al. (2022), describes that there was an incompatibility of leadership between speech and actions, and, although there were attempts of direct channels of communication between the president and the population, his discourse provided relief to only part of the beneficiaries. Drawing from Couto et al. (2022) findings, we noticed that the conflicts of the former president with governors was not present in the communication with the public bank of Caixa, Pedro Guimarães, even though they were both connected and showed an aligned discourse. It suggests that the purported neutrality of the administrative leader did not work even for left-wing voters.

Other questions and limitations remain. First, along the pandemic, there was a “war of narratives” (Couto et al., 2022) in the communication process by leaders. As misinformation has been a huge problem in the last few governments, the content of the message can play a crucial role in shaping and strengthening public confidence, and it is therefore important for leaders to ensure that they effectively communicate and disseminate information in a manner that inspires credibility among the public. In a crisis context, the efficacy of a public leader's discourse alone may be comparatively weaker than that of other tangible actions, such as the issuance of a payment schedule.

An important theoretical limitation regards the construction of stigma as part of the “Brazilian” administrative burden. Stigma depends on cultural context (Madsen et al., 2022) and the concept was originally developed thinking about the US welfare programs (Herd and Moynihan, 2018). Particularly, it can be different in the Brazilian context, but we need to remember the strong associations of beneficiaries of Bolsa Família (Kawauchi, 2019) and the frequent conservative associations between welfare, deservingness, and moral judgements about cash transfers to vulnerable people in Brazil.

In terms of methodology, an important limitation regards the ideology of the leaders. Thinking about the real-world scenario of a right-wing government, we have just manipulated the ideology of the leader to the right, which limited our approach. One could think of replicating the design, but with right and left spectrums.

Moreover, the overrepresentation of southeast participants and their adherence to social networks may have limited our conclusions. Using Facebook as recruitment tool and our sample size might have imposed their own limitations, such as the restriction of the findings to the analyzed participants. However, as we have systematically recruited participants in several groups organized to discuss the “auxílio emergencial”, our findings can be considered for theoretical and practical purposes. Additionally, the process of requesting and receiving the emergency aid was completely digitized since its conception (Marins et al., 2021), including the use of data from Dataprev about potential beneficiaries and the remote action of Caixa Econômica Federal, which opened digital accounts for beneficiaries who did not have a bank account. Therefore, the online means of selecting participants also reproduced this interaction.

As suggestion for future studies, one could think of developing a qualitative and observational study in the Facebook groups to capture other dimensions of burdens. For example, one could observe posts with the presence of congressmen who actively engaged on the debate about the benefits, who demonstrated a high level of influence on social media. Also, due to recent scandals of abuse of the former Caixa president, the impact of the dark side of public leadership (Backhaus and Vogel, 2022) on outcomes like citizens’ trust can be a particularly interesting for research on citizen-state interactions.

References

Aarøe, L., Baekgaard, M., Christensen, J., & Moynihan, D. P. (2021). Personality and Public Administration: Policymaker Tolerance of Administrative Burdens in Welfare Services. Public Administration Review, 81(4), 652–663. https://doi.org/10.1111/puar.13381

Backhaus, L., & Vogel, R. (2022). Leadership in the public sector: A meta‐analysis of styles, outcomes, contexts, and methods. Public Administration Review, 82(6), 986–1003. https://doi.org/10.1111/puar.13516

Baum, M. A., & Groeling, T. (2009). Shot by the messenger: Partisan cues and public opinion regarding national security and war. Political Behavior, 31, 157-186.

Behn, R. D. (1998). What right do public managers have to lead?. Public Administration Review, 58(3): 209-224.

Bellé, N. (2013). Leading to make a difference: a field experiment on the performance effects of transformational leadership, perceived social impact, and public service motivation. Journal of Public Administration Research and Theory, 24, 109-136.

Burden, B. C., Canon, D. T., Mayer, K. R., & Moynihan, D. P. (2012). The Effect of Administrative Burden on Bureaucratic Perception of Policies: Evidence from Election Administration. Public Administration Review, 72(5), 741–751. https://doi.org/10.1111/j.1540-6210.2012.02600.x

Cardoso, B. B. (2020). The Implementation of Emergency Aid as an exceptional measure of social protection. Revista de Administração Pública, 54(4), 1052–1063. https://doi.org/10.1590/0034-761220200267x

Couto, F., Farias Alves Correia, G., & Carrieri, A. de P. (2022). O Antilíder: Da liderança discursiva presidencial à descoordenação federativa para o combate à Covid-19. Cadernos Gestão Pública e Cidadania, 27(87), 1–19. https://doi.org/10.12660/cgpc.v27n87.83879

Crayne, M. P., & Medeiros, K. E. (2021). Making sense of crisis: Charismatic, ideological, and pragmatic leadership in response to COVID-19. American Psychologist, 76(3), 462–474. https://doi.org/10.1037/amp0000715

Goenka, S., & Thomas, M. (2022). Are Conservatives Less Likely Than Liberals to Accept Welfare? The Psychology of Welfare Politics. Journal of the Association for Consumer Research, 7(3), 305–315. https://doi.org/10.1086/719586

Halling, A., & Bækgaard, M. (2022). Administrative Burden in Citizen-State Interactions: A Systematic Literature Review. https://doi.org/10.31219/osf.io/26xdj

Hattke, F., Hensel, D., & Kalucza, J. (2020). Emotional Responses to Bureaucratic Red Tape. Public Administration Review, 80(1), 53–63. https://doi.org/10.1111/puar.13116

Herd, P., & Moynihan, D. P. (2018). Administrative Burden: Policymaking by Other Means (1 edition). Russell Sage Foundation.

Jost, J. T. (2017). Ideological Asymmetries and the Essence of Political Psychology. Political Psychology, 38(2), 167–208. https://doi.org/10.1111/pops.12407

Jost, J. T., & Sterling, J. (2021). The language of politics: Ideological differences in congressional communication on social media and the floor of Congress. Social Influence. 15(2-4), 80-103. https://www.tandfonline.com/doi/full/10.1080/15534510.2020.1871403

Kawauchi, M. (2019). O Programa Bolsa Família e percepções de não beneficiários: Um estudo sobre o efeito preconceito [Universidade de Brasília]. https://repositorio.unb.br/handle/10482/35688

Lagowska, U., Sobral, F., & Furtado, L. M. G. P. (2020). Leadership under Crises: A Research Agenda for the Post-COVID-19 Era. BAR - Brazilian Administration Review, 17(2), e200062. https://doi.org/10.1590/1807-7692bar2020200062

Lau, V. W., Tse, D. C. K., Bligh, M. C., Hong, Y., Kakarika, M., Chan, H., & Chiu, C. P. Y. (2022). Not “My” crisis: Social identity and followers’ crisis responses to COVID‐19. Analyses of Social Issues and Public Policy, 10.1111/asap.12316. https://doi.org/10.1111/asap.12316

Lees-Marshment, J., & Smolović Jones, O. (2018). Being more with less: Exploring the flexible political leadership identities of government ministers. Leadership, 14(4), 460–482. https://doi.org/10.1177/1742715016687815

Lima, K. (2022). Dança das cadeiras: Governo Bolsonaro já acumula quase 30 trocas de ministros desde 2019. G1. https://datawrapper.dwcdn.net/0jrmv/3/

Lindgren, I., Madsen, C. Ø., Hofmann, S., & Melin, U. (2019). Close encounters of the digital kind: A research agenda for the digitalization of public services. Government Information Quarterly, 36(3), 427–436. https://doi.org/10.1016/j.giq.2019.03.002

Lissardy, G. (2020). Trump x Ciência: Os embates incomuns entre o presidente dos EUA e especialistas em clima e coronavírus. BBC News Brasil. https://www.bbc.com/portuguese/internacional-54215423

Madsen, C. Ø., Lindgren, I., & Melin, U. (2022). The accidental caseworker – How digital self-service influences citizens’ administrative burden. Government Information Quarterly, 39(1), 101653. https://doi.org/10.1016/j.giq.2021.101653

Maphinda, H. K. (2017). Leadership and sensemaking: a conceptual analysis (Master’s thesis) Retrieved from http://scholar.sun.ac.za/handle/10019.1/101042. Stellenbosch University, South Africa.

Marins, M. T., Rodrigues, M. N., Silva, J. M. L. da, Silva, K. C. M. da, & Carvalho, P. L. (2021). Auxílio Emergencial em tempos de pandemia. Sociedade e Estado, 36, 669–692. https://doi.org/10.1590/s0102-6992-202136020013

Mau, T. A., Callahan, R. F., & Ohemeng, F. (2022). Guest editorial: Leadership in times of crisis: the intersection of political and administrative leadership. International Journal of Public Leadership, 18(2), 93–100. https://doi.org/10.1108/IJPL-05-2022-118

McLellan, A. (2022). Leadership in a time of crisis: Reflections of a minister in the government of Canada during 9/11 and SARS. International Journal of Public Leadership, 18(2), 101–108. https://doi.org/10.1108/IJPL-02-2022-0008

Moynihan, D., Herd, P., & Harvey, H. (2015). Administrative Burden: Learning, Psychological, and Compliance Costs in Citizen-State Interactions. Journal of Public Administration Research and Theory, 25(1), 43–69. https://doi.org/10.1093/jopart/muu009

Moynihan, D. P., Wright, B. E., & Pandey, S. K. (2012). Working Within Constraints: Can Transformational Leaders Alter the Experience of Red Tape? International Public Management Journal, 15(3), 315–336. https://doi.org/10.1080/10967494.2012.725318

Ospina, S. M. (2016). Collective leadership and context in Public Administration: bridging public leadership research and leadership studies. Public Administration Review, 77(2): 275-287.

Peci, A., Avellaneda, C. N., & Suzuki, K. (2021). Governmental responses to COVID-19 Pandemic. Revista de Administração Pública, 55, 1-11.

Peeters, R. (2020). The Political Economy of Administrative Burdens: A Theoretical Framework for Analyzing the Organizational Origins of Administrative Burdens. Administration & Society, 52(4), 566–592. https://doi.org/10.1177/0095399719854367

Peeters, R. (2023). Digital Administrative Burdens: An Agenda for Analyzing the Citizen Experience of Digital Bureaucratic Encounters. Perspectives on Public Management and Governance, gvac024. https://doi.org/10.1093/ppmgov/gvac024

Peeters, R., & Widlak, A. (2018). The digital cage: Administrative exclusion through information architecture – The case of the Dutch civil registry’s master data management system. Government Information Quarterly, 35(2), 175–183. https://doi.org/10.1016/j.giq.2018.02.003

Samuels, D. J., & Zucco, C. (2012). Using Facebook as a Subject Recruitment Tool for Survey-Experimental Research. https://doi.org/10.2139/ssrn.2101458

Schymura, L. G. (2020, Abril). A dificuldade de o auxílio emergencial chegar a quem precisa [Carta do IBRE]. https://portalibre.fgv.br/sites/default/files/2020-05/cartaibre.pdf

Sobral, F., Carvalho, J., Łagowska, U., Furtado, L. M. G. P., & Grobman, M. (2020). Better safe than sorry: Leadership sensemaking in the time of COVID-19. Revista de Administração Pública, 54(4), 758–781. https://doi.org/10.1590/0034-761220200262x

Thomsen, M. K., Baekgaard, M., & Jensen, U. T. (2020). The Psychological Costs of Citizen Coproduction. Journal of Public Administration Research and Theory, 30(4), 656–673. https://doi.org/10.1093/jopart/muaa001

Tummers, L., Weske, U., Bouwman, R., & Grimmelikhuijsen, S. (2015). The Impact of Red Tape on Citizen Satisfaction: An Experimental Study. International Public Management Journal, 19(3), 320–341. https://doi.org/10.1080/10967494.2015.1027800

Van Wart, M. (2013). Administrative Leadership Theory: A Reassessment after 10 Years. Public Administration, 91(3), 521–543. https://doi.org/10.1111/padm.12017

IR