What Determines the Public’s Evaluations of State-Level Environmental Agencies?

Trust in State Bureaucracy

Americans’ evaluations of state-level bureaucratic institutions are likely shaped by their general levels of approval of and trust in state government. When Americans are asked about their preference for state or national government, they are more likely to embrace state government (see, e.g., Hibbing and Theiss-Morse 2002). And, although trust in state government varies over time and by the state, surveys routinely show that Americans place more trust in state governments than the national government (Weinschenk and Helpap 2015; Wolak and Kelleher Palus 2010).

Wolak (2020) discusses several potential explanations for why Americans may prefer state governments more. First, state and local governments are closer to the people and smaller in size and reach. As Wolak (2020) points out, this could align with Americans’ preference for small government. Additionally, it may be easier for states to provide better representation because states serve smaller, more homogeneous groups of constituents than the national government (Wolak 2020).

These impressions of state governments are important because even though Americans often know less about their state governments (see Bachner and Ginsberg 2021; Lyons, Jaegar, and Wolak 2013; Songer 1984), findings about Americans’ knowledge of state and local politics suggest they matter. For instance, Shaker (2012) finds that Americans who are not as interested in national politics tend to be more focused on state and local politics, even though that would not be readily apparent by looking at overall levels of state and local political knowledge. Additionally, Lyons, Jaegar, and Wolak (2013) argue that while the public may not always be knowledgeable about state politics, they are still active in “monitor[ing] the business of the states” particularly when governments are underperforming, political competition is high, and government is divided (p. 183). Even with limited knowledge, Americans are able to have meaningful evaluations of state institutions.

While overall opinions about state government may be higher in comparison to the federal government, individuals’ impressions of state institutions vary based on the type of institution in question. For example, Hibbing and Theiss-Morse (2002) find that the public varies in its preference for how government should function—that is, who or what institutions should have power and what processes they should use to execute that power. Although bureaucratic institutions are more effective and efficient than the public perceives (Meier and Bohte 2006), people distrust programs that are managed by the government, suspecting government agencies of wasteful spending (Blendon and Benson 2001; Carver et al. 2003). Additionally, the public charges that bureaucracies impose unnecessary rules, paperwork, and interactions, from which neither Americans nor bureaucrats gain any satisfaction.

Political elites help to motivate and continue a negative dialogue surrounding bureaucracy. As Garrett et al. (2006) point out, candidates for political office frequently “bash” the bureaucracy as part of campaign rhetoric, “pledging to cut fraud, waste, abuse, and excessive regulation and reduce bureaucratic intrusiveness” (p. 228). These cues, such as the mere word “bureaucracy,” are argued to be “essential in maintaining electoral coalitions” (Fishel 1985, 10; Garrett et al. 2006). This sort of bashing makes sense from a political standpoint, since it places blame onto those who cannot be electorally punished, but it does incur another kind of punishment: distrust toward government agencies (Garrett et al. 2006). According to government agency managers interviewed by Garrett et al. (2006), the public, who already lacks considerable knowledge about the inner-workings of the “black-box” bureaucracy, fills in the gaps in their knowledge with incomplete and negative information.

In sum, while Americans tend to have a more positive view of state institutions, their trust in and approval of state bureaucracy varies. I expect that this variation in trust in state bureaucratic institutions has an effect on how the public views individual agencies. In particular, I expect that as individuals’ trust in state bureaucracy increases, we should see more positive evaluations of state-level environmental agencies.

Partisan Lens

While individuals may express some skepticism about state bureaucracy generally, they can still hold positive opinions about individual agencies, agency interactions, and/or government programs run by unelected experts (Van de Walle, Kampen, and Bouckaert 2005; see also Hibbing and Theiss-Morse 2002). For example, while voters who align with the Democratic Party have historically supported social programs and the agencies that implement them, they are often more suspicious of agencies dedicated to military and police power (Hall 2002). Furthermore, whereas the public may prefer unelected experts to elected officials (see Hibbing and Theiss-Morse 2002) or prefer unelected experts in a highly technical policy area, such as energy policy, (see Ruostetsaari 2010), they may be reluctant to hand over power to unelected experts over policy and implementation that is more perceivably subjective, such as abortion. Researchers have even been able to shift opinions about specific agencies by speaking about the value of particular agencies using various ideological frames—that is, speaking to the ideological preferences of individuals in a way that makes them reflect more positively on the agency in question (Anson and Kane 2023). What these findings about individuals’ agency and program assessments have in common is that opinions appear to be shaped by partisan or ideological preferences.

As noted above, knowledge of state politics and institutions is limited and may be salient only under certain conditions. Because agency workers are not politically identifiable, the public’s opinions about bureaucratic agencies can depend upon political officials’ accounts of the successes/failures of agencies (Darmofal 2005). In evaluating institutions and elected officials, partisanship often “serves as a powerful ‘perceptual screen’” (Bisgaard 2015, see also Campbell et al. 1960, 133). Studies focused on explaining how voters connect their perceptions of the economy to their support of elected officials often find that voters’ partisan identification shapes which party/elected official they credit or blame for economic conditions (Bisgaard 2015; see also Bartels 2002; Evans and Andersen 2006; Wlezien, Franklin, and Twiggs 1997). If my party is in power, I may be less likely to see poor economic conditions as attributable to the current government’s actions. Therefore, to overcome a lack of knowledge about state-level environmental agencies, Americans likely rely on cues, such as ideological cues and cues from political elites, to form their opinions about these environmental agencies.

However, the strength with which partisan identification shapes the assignment of responsibility for policy conditions is contingent on several factors, including the ability of the public to easily attribute responsibility (Bisgaard 2015) and the clarity of policy conditions (Parker-Stephen 2013). In easily attributing responsibility, federalism presents a problem. In order for the public to accurately assign credit/blame (without relying mostly on partisan cues to do so), they need to understand the responsibilities of local, state, and federal governments (i.e., what each level of government controls and what each level can and cannot do). Unfortunately, it appears that federalism and other complex institutional features, such as divided government, “blur lines of responsibility and make it difficult for voters to hold governments to account” (Tilley and Hobolt 2011, 316; see also Anderson 2000; Hellwig and Samuels 2008; Powell and Whitten 1993; Whitten and Palmer 1999). Because of federalism, we can expect that partisanship may play a larger role in shaping evaluations of state-level environmental agencies than other factors.

In determining how partisanship may affect evaluations of state-level environmental agencies, we can look to the divisions between the parties in regards to environmental policy issues. According to McCright and Dunlap (2011), there are marked differences between how Democrats and Republicans view important environmental problems, such as climate change. As they note, “environmental protection typically entails governmental intervention into markets and restrictions on property rights, challenging conservative values” (p. 160). Democrats are more likely to be receptive to regulatory action to protect the environment than Republicans. Environmental protection agencies tend to rely on regulating industry as a primary means of protecting the environment (see Hopper 2020). Thus, it is not surprising that only around 25% of Republicans state that the federal EPA is doing “an excellent or good job,” in comparison to Democrats’ 54% (Jones 2021). Furthermore, we might expect that this sentiment would translate to the state level, where the public may rely even more heavily on partisan cues due to their limited knowledge of state institutions and the complexities introduced by federalism.

Thus, in assessing how ideology or partisanship may affect opinions on state-level environmental agencies, it is likely that two factors may matter: (1) how an individual’s partisanship and strength of that partisanship may shape their impressions of a regulatory environmental agency and (2) how an individual’s shared or opposing partisanship with the current government may shape who they look to for cues and their understanding of the success of agencies managed by that current government. I expect that the stronger that individuals identify with the Republican Party, the more negative their impressions of state-level environmental agencies will be.

I also expect that when an individual’s partisanship matches that of the current government, they will be more positive about agencies under that government’s control, including state-level environmental agencies. As to which partisan match is most relevant, governors across the states have the ability to oversee and manage bureaucratic agencies. Additionally, Americans often hold executives accountable for agency failures rather than their individual legislators (see Malhorta and Kuo 2008, for example). Therefore, I expect that when individuals’ partisanship matches the partisanship of the governor, their impressions of state-level environmental agencies will be more positive.

Environmental Policy Outcomes

While partisanship likely provides a useful cue for the public in assessing state-level environmental agencies, environmental policy outcomes may also have a meaningful effect. For example, Yackee and Lowery (2005), find that public approval of the bureaucracy “is, in part, controllable by public managers with regards to scandals, policy failures and budget management. . .” (p. 531). Research across policy areas supports that the public’s impression of agency actions and policy outcomes may affect their impression of the agency managing the policy area in question. Shaw et al. (1998) find that impressions of crime and safety affect public support for community police. Support for NASA and its programs declined after both the Challenger and Columbia disasters (Roper Center 2015). And, views of Department of Defense programs and the military shift based on Americans’ perceptions of ongoing military conflicts (Gallup “Military and National Defense”).

As it relates to environmental policy, environmental disasters and poor environmental quality have often served as the catalyst for large-scale changes in environmental enforcement, including the creation of the federal EPA and the passage of significant environmental laws, such as the Clean Air Act and the Clean Water Act. Although specific assessments of evaluations of environmental quality and those evaluations’ implications for opinions of agencies are rare, surveys of Chinese citizens show that perceptions of environmental pollution had significant and positive effects on citizens’ satisfaction with environmental protection efforts (Ruan et al. 2022).

One complication in connecting environmental policy conditions and assessments of environmental agencies is that the public may not have a clear understanding of policy conditions. While some indicators are fairly straightforward and apparent (e.g., unemployment rates), others may be more difficult for the public to grasp. For example, while air quality has improved considerably by some metrics, surveys find that Americans actually believe air quality to be worsening (Graves 2003). This does not mean that Americans are wrong, however. Americans likely base their assessments of the environment on local conditions (as opposed to national levels of pollution), and their assessments may be more reliant on visible forms of air pollution, such as smog. Moreover, as the economic voting literature finds, assessments of policy outcomes do not have to be accurate for those assessments to be leveraged against institutions and elected officials.

In sum, while partisanship is likely a powerful factor in shaping assessments of individual agencies like state-level environmental agencies, it is likely that policy outcomes also affect the public’s perceptions of agencies. Therefore, I expect that positive impressions of states’ environmental quality should lead to more positive assessments of state-level environmental agencies.

Data and Methods

To assess my hypotheses, I rely on data from a 2020 national survey, in which I asked respondents to answer questions regarding their trust in and preference for state governments and bureaucracy, their assessment of their state’s environmental quality, their assessment of their state-level environmental agencies’ performance, and several other questions related to political interest, media usage, party identification and strength, and demographics.

Variables

All variables except for a measure of state environmentalism are derived from the 2020 survey. The dependent variable in the analyses that follow is the assessment of the success of environmental agencies. To measure this, I rely on a survey question asking respondents to assess the success of their relative environmental agencies in controlling pollution (see above). Since the primary independent variable is an assessment of environmental quality, success is specified as the ability to control pollution (as opposed to asking about success generally). While measures of assessment exist for federal-level agencies and of government or layers of government, generally—to my knowledge, there are no existing measures of the public’s evaluation of their relative state-level environmental agencies. Thus, the survey findings provide us with a direct measure of assessment.

The primary independent variables for this analysis are respondents’ trust in state bureaucracy, their assessment of environmental quality, their party identification and the strength of that identification, and whether or not the party identification of respondents matches the party of the governor in power. First, in line with research on the factors that shape the evaluation of institutions, I include a measure of trust in state bureaucracy. This measure is derived from a survey question that asks respondents on a scale from strongly disagree to strongly agree whether they trust their state’s bureaucratic institutions.

To measure the public’s evaluation of environmental policy conditions, I include a variable reflecting respondents’ assessment of environmental quality in their respective states. This reflection is in response to a question prompting respondents to evaluate the quality of their respective state’s environment. In addition to environmental policy conditions, I also account for respondents’ party identification and whether or not their partisan identification matches the party in control of the governorship. To measure party identification and the strength of that identification, I use a 7-point party identification scale, in which weak Democrats are a one and strong Republicans are a seven. ¹ To identify a partisan match between governors and respondents, I determined whether or not the party of the governor in power matched the partisan identification of the respondent, using a 3-point party scale. A match (e.g., Democratic respondent with Democratic governor) is considered a one, while a mismatch (e.g., Democratic respondent with Republican governor or Independent respondent with Republican/Democratic governor) is considered a zero. ²

In addition to the independent variables of interest, I also include control variables to account for demographics (e.g., race, income, education, and gender) and interest in politics and news. In particular, survey questions that asked respondents about their interest in state politics and news can help to give us a better sense of how aware they may be of policy conditions and state institutions, generally. I have also included a measure of environmentalism for the respondents’ respective states. To measure environmentalism, I rely on the League of Conservation Voter (LCV) average scores for each respective state’s 2019 House delegation. While LCV scores do not reflect the attitudes of the public directly, they offer a proxy measure of environmentalism by reflecting the preferences of the public on environmental policy issues via the public’s elected representatives. LCV scores are commonly used in state-level environmental policy research for this purpose (Konisky and Woods 2012; see also Hopper 2017). Table 1 lists the summary statistics and more detailed information about the measurement of the variables used in this analysis.

OPEN IN VIEWER

Table 1.

Summary Statistics.

Variable	Observations	Mean	Std. deviation	Min	Max
Environmental agency success	1,048	3.51	1.09	1 (strongly disagree)	5 (strongly agree)
Assessment of environmental quality	1,048	3.46	0.99	1 (very low quality)	5 (very high quality)
Trust in state bureaucracy	1,048	1.97	0.65	1 (not at all)	3 (most of the time)
Seven-point party identification scale	1,048	4.20	1.97	1 (strong Democrats)	7 (strong Republicans)
Governor match	1,048	0.49	0.50	0 (match)	1 (mismatch)
Interest in state politics	1,048	3.00	0.95	1 (not at all interested)	4 (extremely interested)
Interest in news	1,048	3.15	0.90	1 (not at all interested)	4 (extremely interested)
Years living in the state of residence	1,048	3.43	0.89	1 (less than a year)	4 (10+years)
Non-white	1,048	0.22	0.42	0 (white)	1 (non-white)
Income	1,048	6.39	3.71	1 (less than $10,000)	12 (more than $150,000)
Highest degree of education	1,048	3.94	1.69	1 (less than high school degree)	8 (doctoral degree)
Male	1,048	0.49	0.50	0 (not male)	1 (male)
Female	1,048	0.50	0.50	0 (not female)	1 (female)
League of conservation voters state average	1,048	56.21	25.83	3 (low score)	100 (high score)

To test my hypotheses, I use a generalized least squares (GLS) model, with errors clustered by state. As the nature of the data and diagnostics suggest that ordinary least square would be inefficient, GLS may be used to account for heteroscedasticity in the errors (Wooldridge 2009, 278). ³ Additionally, the use of clustered standard errors allows me to control for the likelihood of some unique state-level effects.

Results

Figures 1 and 2 provide us with some insight about the distribution of the survey data for the two most unique variables in the dataset: the respondents’ evaluations of their respective state-level environmental agencies and respondents’ evaluations of their respective states’ environmental quality. Looking at Figure 1, we can see that a majority of respondents (53.72%) somewhat or strongly agree that their state’s environmental agency effectively controls pollution, while around 30% of respondents neither agree or disagree that their state’s environmental agency effectively controls pollution. Fewer respondents strongly or somewhat disagree that their state’s environmental agencies are effective. In general, respondents are either inconclusive or mostly positive in assessing the effectiveness of their state environmental agencies.

OPEN IN VIEWER

Figure 1.

Every state has an environmental agency that focuses on controlling pollution. How much do you agree with the following statement: “my state’s environmental agency is successful in controlling pollution”?

OPEN IN VIEWER

Figure 2.

What is your impression of the quality of your state’s environment (i.e., clean air and water)?

In Figure 2, we can see a similar trend when we look at respondents’ evaluation of environmental quality in their respective states. Around 46% of respondents rate the quality of their state’s environment as either high or very high quality, while almost 40% of respondents rate the quality of their state’s environment as average. As was the case with environmental agency evaluations, respondents appear relatively positive in their evaluations; however, there is a 7.25% difference between the majority of respondents that evaluate their environmental agencies highly and those that rank their state’s environmental quality highly. While this difference is notable, the trend in responses still suggests that these evaluations of quality may be connected to assessments of the effectiveness of state-level agencies.

To get a clearer picture of the predictors of these assessments of effectiveness, we can turn to the model in Table 2. The results displayed in Table 2 suggest that respondents’ assessments of their respective state’s environmental quality and respondents’ trust in state-level bureaucratic institutions are strong predictors of their assessments of the effectiveness of state-level environmental agencies. These findings support H1 and H4, in which I expect that higher levels of trust in state bureaucratic institutions and higher environmental quality assessments will lead to increased positivity in assessments of state environmental agency effectiveness. In evaluating partisan effects (H2 and H3), we cannot confidently establish with this data and model that either the strength of partisan identification or a partisan match with the governor in power are predictors of state-level environmental agency assessments. Similarly, other notable control variables that we might expect to matter, such as interest in state politics, interest in the news, state-level environmentalism, and demographics are not significant in the model. The model in Table 2 is driven primarily by trust in state bureaucracy and state-level environmental quality evaluations.

OPEN IN VIEWER

Table 2.

Effects of Environmental Assessments, Trust in State Bureaucracy, and Partisanship on Evaluations of State-level Environmental Agencies.

Variables	Coefficients
Assessment of environmental quality	0.65** (0.04)
Trust in state bureaucracy	0.27** (0.05)
Seven-point party identification scale	0.01 (0.01)
Governor match	0.03 (0.05)
Interest in state politics	0.05 (0.04)
Interest in news	−0.02 (0.04)
Years living in the state of residence	0.01 (0.02)
Highest degree of education	0.02 (0.02)
Income	−0.01 (0.01)
Non-white	0.05 (0.07)
Male	0.14 (0.33)
Female	−0.01 (0.32)
League of conservation voters state average	−0.01 (0.01)
Constant	0.44 (0.37)

Note: Beta coefficients are listed, with standard errors in parentheses. Standard errors are clustered by state.

n = 1,048 | R² = 0.53

**

p < .01.

Given that the literature suggests partisan evaluations are likely to matter in the assessments of state-level institutions, it is worth looking more closely at the variables in the model to see whether they are entirely distinguishable from one another. It is possible, for example, that assessments of environmental quality and/or trust in state bureaucracy are correlated with partisanship. The literature I note above expects that these opinions may be driven by partisan cues, so a correlation between the variables that diminishes the effects of partisanship may explain the lack of significance in the model.

To assess this possibility, we can, first, look at Pearson’s correlation statistics in Table 3 for environmental quality evaluations, trust in state bureaucracy, strength of partisanship, and a partisan match with the governor.

OPEN IN VIEWER

Table 3.

Correlations Between Environmental Conditions, Trust in Bureaucracy, and Partisan Variables.

	Assessment of environmental quality	Trust in state bureaucracy
Seven-point party identification scale	0.13*	0.01
Governor match	0.11*	0.11*

*

p < .01.

While the strength of the correlations between the variables is relatively low, we can be confident that assessments of environmental quality are correlated with strength of partisanship and a partisan match with the governor and that trust in bureaucracy is correlated with a partisan match with the governor. These correlations suggest that it is possible that that these variables may measure similar dynamics, causing our confidence in the partisan effects to weaken when we include evaluations of environmental quality and trust in state bureaucratic institutions.

Indeed, looking at Table 4, which provides a base model that includes only the partisan predictors, we can see that H3 is supported, with a partisan match with the governor leading to stronger evaluations of environmental agency effectiveness. Also, in the absence of the trust in state bureaucracy and environmental quality evaluation variables, other notable controls become significant, including interest in state politics and income levels. However, H2 is not supported. While we can be confident using the data and model in Table 4 that respondents’ strength of partisan identification likely shapes (although, not strongly) their assessments of state environmental agency effectiveness, the direction of the effect is the opposite of what I expect in H2. Instead, the model suggests that stronger identification with the Republican party leads to an increase in the positivity of assessments of environmental protection state environmental agencies. As a reminder, this effect drops from significance when the variable is included in the full model in Table 2.

OPEN IN VIEWER

Table 4.

Model of State Environmental Agency Assessments without Environmental Conditions or Trust in Bureaucracy Variables.

Variables	Coefficients
Seven-point party identification scale	0.05*** (0.01)
Governor match	0.19*** (0.05)
Interest in state politics	0.20*** (0.06)
Interest in news	0.02 (0.05)
Years living in the state of residence	−0.02 (0.03)
Highest degree of education	0.06* (0.03)
Income	0.02** (0.01)
Non-white	0.11 (0.09)
Male	0.98 (0.75)
Female	0.59 (0.74)
League of Conservation Voters state average	0.01 (0.01)
Constant	1.29 (0.79)

Note: Beta coefficients are listed, with standard errors in parentheses. Standard errors are clustered by state.

n = 1,048 | R² = 0.17

***

p < .01. **p < .05. *p < .10.

Table 5 includes both models so that we can easily compare coefficients and the R² statistics for the models.

OPEN IN VIEWER

Table 5.

Comparing Models One and Two.

	Model 1	Model 2
Assessment of environmental quality	0.65*** (0.04)	—
Trust in state bureaucracy	0.27*** (0.05)	—
Seven-point party identification scale	0.01 (0.01)	0.05*** (0.01)
Governor match	0.03 (0.05)	0.19*** (0.05)
Interest in state politics	0.05 (0.04)	0.20*** (0.06)
Interest in news	−0.02 (0.04)	0.02 (0.05)
Years living in the state of residence	0.01 (0.02)	−0.02 (0.03)
Highest degree of education	0.02 (0.02)	0.06* (0.03)
Income	−0.01 (0.01)	0.02** (0.01)
Non-white	0.05 (0.07)	0.11 (0.09)
Male	0.14 (0.33)	0.98 (0.75)
Female	−0.01 (0.32)	0.59 (0.74)
League of Conservation Voters state average	−0.01 (0.01)	0.01 (0.01)
Constant	0.44 (0.37)	1.29 (0.79)
	R2 = 0.46	R2 = 0.17

Note: Beta coefficients are listed, with robust standard errors in parentheses.

n = 1,048.

***

p < .01. **p < .05. *p < .10.

In comparing the two models, it is clear that while the variables may be correlated, when they are all included together, we can be more confident that respondents’ assessments of environmental quality and respondents’ trust in state bureaucracy are powerful predictors of environmental agency evaluations than we can be confident in the purely partisan variables. Even with controlling for partisan factors in Model 1, the evaluations of environmental quality and trust in state bureaucracy variables boast substantively stronger coefficients than the partisan variables, alone, in Model 2. Furthermore, the significantly higher R² in Model 1 (0.53 compared to 0.17) suggests that the model including environmental quality evaluations and trust in state bureaucracy more adequately explains the variation in attitudes around state-level environmental agencies than the model that includes only the partisan variables. In sum, there is a probable connection between respondents’ partisanship and their evaluations of environmental quality and trust in state bureaucracy. However, evaluations and trust are likely helping to explain something about respondents’ assessments of environmental agencies that partisanship, alone, cannot.

Discussion and Conclusions

Although we have some understanding of the way Americans perceive of the national government’s handling of environmental issues, we know far less about the public’s impression of state-level efforts. This is particularly the case when it comes to our understanding of the public’s impressions of state-level environmental agencies—agencies that play a significant, if not dominant, role in American environmental enforcement efforts (Hopper 2020). The bureaucrats working in state-level environmental agencies are not immune to public opinion. While workers in state environmental agencies are indirectly accountable to the public via the oversight and control of elected officials, they still feel the direct pressures of public opinion and the effects their actions may have on local economies—economies that they, their families, and their neighbors live and work in (Konisky 2008; Koontz 2002; see also, Hopper 2017). Successful bureaucratic institutions rely on legitimacy invested in them by the public and elected officials in order to successfully pursue their mandates (Carpenter 2010; Rogowski 2020; see also Kim 2005; Levi 1997; Tyler 2006; Yates 1982). Thus, understanding the factors that shape the public’s impression of institutions is crucial.

The findings, here, suggest that several factors play into how Americans evaluate their state-level environmental agencies. Respondents’ evaluations of environmental quality in their states and respondents’ trust in state bureaucratic institutions appear to have a substantive effect on evaluations of state-level environmental agencies. This supports existing research that finds that individuals’ feelings about state government, bureaucratic institutions, and their consideration of policy conditions and context affect institutional assessments. My findings also support that the factors that drive evaluations of these environmental agencies are difficult to untangle from partisanship and ideology. Both assessments of environmental quality and trust in state bureaucratic institutions are correlated with political factors. As I note previously, the public often relies on partisan cues and may use partisanship as a lens through which they perceive policy conditions and institutions. It is not surprising, then, that we see partisanship so closely aligned with other opinions. And, yet, even when those partisan factors are considered, the effects of policy considerations and trust in state bureaucracy are powerful and provide a more complete picture of the variety of factors that may be shaping opinions about state environmental agencies.

This being said, the models I present, here, have several limitations. First, we are looking at only a single snapshot in time, as I rely entirely on data collected from the 2020 survey. ⁴ More data are necessary to come to firm conclusions. Additionally, known issues with online survey tools like Qualtrics should be considered when interpreting the results of these analyses. Again, a variety of survey research on this topic over a longer period of time would provide more confidence in the conclusions offered here.

Another important consideration is in the ordering and wording of survey questions. The ordering of the questions may have boosted the effect of the evaluations of environmental quality on assessments of environmental agencies. In the survey, respondents were asked about environmental quality assessments directly prior to being asked about their assessments of environmental agencies. This may have led them to connect the two responses, with environmental quality assessments at the top of their mind, even if that connection would not have been relevant outside of the context of the survey. Also, while pilot surveys at the University of Tampa in 2019 did not prompt concerns about question wording, it is important to remain aware of how the questions were posed and what that might mean for the substance of the measures. Are these questions able to garner what the public truly thinks of state bureaucracy, state environmental agencies, and environmental quality? More research in this area will help to determine the effectiveness of these measurement tools.

Despite limitations, this research offers us one of the first looks at what the American public thinks of its state-level environmental agencies and what factors might be shaping those impressions. Furthermore, the findings, here, contribute to the broad literature exploring public opinion on state-level institutions by providing support for findings that trust/preference, policy conditions, and partisanship often drive public opinion of state institutions, processes, and programs. While many of those state-level studies focus on institutions more generally (i.e., legislatures, the courts, and bureaucracy), this study focuses on a specific type of institution within the category of the state bureaucracy. In a country where states and localities must play a significant role in solving domestic and global environmental problems, it is critical that we understand how the public’s evaluations of state-level environmental agencies may affect their capacity to regulate. This research moves us a step closer to that understanding.

Supplemental Material