The Environmental Protection Agency’s (EPA) enforcement priorities routinely come under criticism for the appearance of inappropriate targeting on the one hand, and the lack of targeting on the other. This study explores programmatic determinants of Federal oversight inspections conducted under the Resource Conservation and Recovery Act (RCRA), including how a state’s performance manifested through inspection intensity affects EPA’s decisions to investigate within a state jurisdiction. The models with year and regional fixed effects suggest that increases in state inspection intensity from the preceding year have a significant effect on reducing the rate of Federal inspections. A theoretical typology for enforcement activity is proposed for future consideration and modeling.
Get Premium Budhwar Peth Call Girls (8005736733) 24x7 Rate 15999 with A/c Roo...
Hart 2014 -- Overseeing the Enforcers (Delivered on Nov. 6, 2014)
1. 1
Overseeing the Enforcers:
Bureaucratic Discretion in Federal
Monitoring of State Hazardous Waste
Enforcement Programs
Nicholas R. Hart
nrhart@gwu.edu | www.nickhart.us | @nickrhart
Trachtenberg School of Public Policy and Public Administration
The George Washington University
November 6, 2014
APPAM Fall Research Conference
Albuquerque, NM
2. Disclaimer
• The views presented in this paper are my own and do
not represent the official position of the U.S.
Government or the U.S. Office of Management and
Budget (OMB).
All information reported in this paper is based on
publicly available documents, reports, and
publications.
N. HART | nrhart@gwu.edu Overseeing the Enforcers 2
3. Background and Overview
• Resource Conservation and Recovery Act
(RCRA) provides for cradle-to-grave
management of hazardous wastes
• 48 of 50 states have accepted some level of
delegated authority
• EPA launched State Review Framework (SRF)
to ensure “consistent” enforcement across
states
N. HART | nrhart@gwu.edu Overseeing the Enforcers 3
7. RCRA Primacy Index, 2012
IA
0
CA
75
AK
0
ID
100
ME
36
KS
46
TX
84
GA
89
N. HART | nrhart@gwu.edu Overseeing the Enforcers 7
8. State Inspection Intensity
Effect on Fed Inspection Intensity
N. HART | nrhart@gwu.edu Overseeing the Enforcers 8
9. Conclusions for OLS/2SLS Models
• Increases in State inspection intensity are
followed by decreases in Federal inspection
intensity
• Evidence of some bureaucratic discretion is
exercised by Federal regional EPA employees
– However, the average effect is relatively small
• Coefficients in the risk, task, and economic
vectors are stable across specifications.
N. HART | nrhart@gwu.edu Overseeing the Enforcers 9
13. Describing the Typology with Inspections
- Violations
- Primacy
- Violations
- TSDF/LQG universe
+ RCRA universe
- emissions/GDP
+ Violations
+ PPG
+ Mfg/GDP
- RCRA universe
+ Violations
+ Area
+ Primacy
- PPG
N. HART | nrhart@gwu.edu Overseeing the Enforcers 13
14. Conclusions on the Typology
• Could be a useful framework for considering
differential treatment-response based on
Federal or state lens.
• More work is needed to refine the typology.
N. HART | nrhart@gwu.edu Overseeing the Enforcers 14
15. Next Steps
• Need to better understand the sensitivity to
types of regulated facilities – TSDFs/LQGs,
SQGs, and CESQGs may receive different
treatment than in the aggregate.
• Explore additional theoretical support for
states in Quadrant II.
N. HART | nrhart@gwu.edu Overseeing the Enforcers 15
Editor's Notes
Thank you to Sara and David.
Project started back in 2012 for a course on politics and public policy as I was starting the PhD program.
Much of the research in this area tends to look at state factors as the dependent variable. In cooperative Federalism, we often want to know how well states do something.
My own experience working in the Federal government, raised questions about how the Federal government responds to state actions.
Wearing a dual hat, I was also taking a course with heavy emphasis on various typologies to think about public policy issues.
This paper was the outcome of my thinking about typologies and how they might be applied to one of the issues that comes up in cooperative Federalism – how do states and the Federal government complement (or not) each other.
Finally – this is very much a work in progress, so I look forward to your helpful feedback, suggestions, and comments.
Mandatory Disclaimer
RCRA is the Federal authority for managing hazardous waste handling.
Nationally there are 800 TSDFs, 23,000 LQGs, 151,000 SQGs, and 196,000 CESQGs.
Much of this authority is delegated to states –
Periodic questions arise about how well states are doing in enforcing the statute – but also how well EPA goes about targeting enforcement activities.
Notably, states are required to minimally enforce RCRA Federal requirements but can always go above any beyond.
In 2004 EPA launched the State Review Framework in collaboration with ECOS.
Intended to be a tool to constantly assess state performance. Reviews are conducted on a four year cycle.
The first round occurred from 2004-2007, the second from 2008-2012, and EPA is currently in the third round.
Think about the interaction within a principal-agent framework – each actor is simultaneously a principal and an agent.
Question – are the level of inspections provided by EPA uniform or is there some variation across states within different regions?
The answer may seem obvious, but to any close observer of EPA policy decisions – and the commentary offered – regions are treated the same by HQ.
For example, if new staff are provided, each region receives one.
Regions are incredibly influential in the EPA apparatus.
What I will present represents two ways of thinking about this issue within a cooperative Federalism context.
The First is a more traditional analysis based on OLS and 2SLS identification strategies.
Then I will develop a typology for thinking about enforcement issues – and descriptively examine quadrants using logits.
Note the two enforcement-related variables of interest –
Federal inspection intensity has a large range but a relatively low mean. This is the number of facilities inspected, not the total number of inspections conducted.
State inspection intensity ranges from zero to 50%. The zero states are Alaska and Iowa. But the mean is just above the Federal maximum.
Also, just want to flag the creation of a metric for cataloging primacy that I have not seen elsewhere. Typically we would code this as dichotomous. But with RCRA we have the ability to look at the proportion of required and voluntary regulations a state has adopted.
This is just another indication of what the primacy index might look like spatially for a single year – there is considerable variation especially in the plains and New England.
This chart presents the key independent variable of interest.
Model 1 is State FE. Not a very strong model overall.
More weight in Region FE models.
Model 2 is just an OLS approach – and the coefficient is significant but relatively small.
Keep in mind though, that the average was relatively small. But this magnitude is about a 4 percent change from the average Federal intensity.
Models 3 & 4 present instrumental variables – to account for potential endogeneity in the joint determination between Federal and State inspections, assuming not all endogeneity is addressed by lagging the state intensity a full year.
While the coefficients are larger in magnitude, the instruments are relatively weak and we can’t conclude these models are better than Model 2 – OLS.
Based on the initial models, we can conclude Federal inspections are sensitive to changes in State inspections.
The effect sizes appear relatively small, but it represents anywhere from a 4 to 14 percent change in Federal inspection intensity.
Conclude that Federal EPA staff are exercising some discretion in targeting Federal resources.
Now a slightly different perspective:
If we think about state and Federal enforcement actions as an X-Y plot – there are potentially different relative perspectives and perspectives that basically agree.
In Quadrant I and III – the principals and agents agree on a relative level of enforcement activities.
Low Risk – relatively small facilities or low history of violations, leading EPA and States to both de-prioritize a level of enforcement.
High Risk – identified problems or potential risk lead the state and EPA to prioritize enforcement.
In Quadrants II and IV though – the principals and the agents disagree.
Quad 4 – states provide a relatively higher level of enforcement than the Federal government thinks is necessary. Perhaps state laws are more stringent than Federal minimum standards.
Quadrant 2 – potentially the most interesting from a Federalism perspective – where the Federal government disagrees with relative enforcement intensity and provides a higher level of inspections. We expect that states that decline primacy would be situated here, but what other states? Perhaps states with incentives to minimally enforce laws, or those faced by resource constraints?
Next – attempt to assign the available state data to quadrants using mean centering (national average not average of states).
Looking at this presents a few potential issues –
Question about where to draw the lines for the axes –
Mean centering can change by year – states may shift to a new quadrant without actually changing their behavior but because other states change theirs.
The observations are largely clustered at about 5 and 0.5 pct. – suggests mean centering may not be the best approach.
Next – will attempt to examine the quadrants in logits to better describe which states are likely to hit each quadrant.
This chart just presents the significant odds ratios to describe each quadrant from four independent logistic regressions.
You can see that there’s still a fair amount of discordance in these models, so this is by no means perfect.
But it does give us some insights in to better articulating how states stack up in the typology.
Take for instance Quadrant 3 – which is where the Federal and State agencies are in relative agreement about inspection intensity relative to the means. There’s a higher probability of being in this quadrant if a state has a large magnitude of violations and is also a PPG recipient but a low probability if there’s a large regulated universe.
Quadrant 2 – is difficult to explain. This descriptive analysis only suggests that there’s a lower probability of sitting here based on RCRA primacy and violations.
DN really get better when removing Alaska or Iowa.
This is another way to look at the same information just summarizing the key factors.
More work is needed.
Perhaps mean centering for the quadrant determinations is inappropriate, particularly based on clustering.
But this may still be a useful tool to think about how states approach enforcement issues.
More work is needed to consider the typology – particularly considering different facility types regulated under RCRA.