Presidential Popularity and Congressional Control of the Bureaucracy: The Clinton Administration and the Contract with America

 

 

 

David M. Hedge

Renée J. Johnson

 

University of Florida

Department of Political Science

 

 

Jeff Gill

 

University of California-Davis

Department of Political Science

 

 

 

 

 

 

 

 

 

 

 

 

Prepared for presentation at the Annual Meeting of the Midwest Political Science Association, April 15-18, 2004, Chicago, IL.

 

 

Introduction

There is now a considerable body of evidence that looks at political control of public bureaucracies, much of it guided by a principal-agent model.  Over the last two decades the focus of that analysis has shifted from asking whether political principals can control their bureaucratic agents (they can and do) to questions concerning the nature of that influence.  Three questions guide much of the recent analysis – What considerations limit the influence of political principals?  What tools do political principals use to control the bureaucracy?  Under what conditions is political control likely to occur?  Taken together, answers to those questions suggest that the process of control is more complex than perhaps the earlier work in this area might suggest.  Three important lessons emerge from that discussion.  First, earlier political control models probably overstated the passivity of administrative agents and the extent to which a hierarchal relationship exists between politicians and bureaucrats.   Research from a number of quarters (e.g. Eisner and Meier, 1990; Krause, 1996; and Golden, 2000) demonstrates that bureaucrats are active participants in bureaucratic politics.  Second, scholars have become more mindful that the actions of one set of political principals cannot be understood in isolation of other, often competing political interests (see, e.g. Moe, 1985; Wood and Waterman, 1994; and Krause, 1999).  While previous research acknowledges the fact of multiple principals, little evidence has been generated to get at the extent to which the actions and attributions of one political principal effects the ability of other principals to exert control.  In methodological terms, prior research tends to assume, implicitly at least, that the influence of multiple principals is additive but not interactive.  Third, recent analysis suggests that while political principals rely on a variety of tools to exert control (e.g. hearings, budgets, appointments or administrative procedures) some are more effective than others (see, e.g. McCubbins, Noll, and Weingast, 1987 and Wood and Waterman, 1994).   Finally, evidence from a number of quarters suggests that efforts at bureaucratic control are shaped by the larger political and economic climate.

Our paper attempts to add to that discussion by looking at congressional control of the bureaucracy during the Clinton administration.  Data from two regulatory agencies – the Occupational Safety and Health Administration (OSHA) and the Nuclear Regulatory Commission (NRC)- are used to examine how presidential standing conditioned the ability of congressional Republicans to reduce the rate of federal regulation following the 1994 election.  A number of questions guide that analysis.  How successful were congressional Republicans in reducing federal regulation? What mechanisms did they use to achieve that end?  Was congressional control of the bureaucracy during the Clinton era conditioned by the President’s standing with the public?  Were federal agencies more likely to respond to congressional pressures when the public’s support for the president was at its lowest? 

To answer those questions, we begin by describing how Republican leaders in the House sought to affect regulatory control during the first days of the Republican “Revolution.”  We then examine how each of our agencies responded to the changing character of congressional and presidential politics.

The Republican takeover of Congress following the 1994 election provides a unique vantage to revisit the issue of political control.  Republican victories in the House and Senate were important from a number of perspectives.  From an institutional perspective, the “Republican Revolution” was not only the first time since the early fifties that the party controlled both houses of Congress, but it held out the possibility of congressional (versus presidential) government as party leaders used the Contract with America to set Congress’ and, for that matter, the nation’s agenda. For regulatory agencies, a Republican Congress represented a more immediate challenge.  Led by Tom Delay, House Republicans mounted an attack on regulation that rivaled the efforts of the Reagan Administration to get the federal government “…off the backs of the American people.”  From an analytic perspective, the Republican takeover offers a relatively rare historical opportunity to examine political control.  By necessity, most of the literature on political control looks at either the impact of changes in presidential administrations or variations in the character (e.g. committee ideology) of what had been, for forty years, a Democratic-controlled Congress.  The 1994 elections offers scholars the chance to examine the impact of a much more infrequent political event – a change in the partisan control of Congress – on the character of federal regulation during a particularly tumultuous period of American politics.

Regulatory Reform in the 104th Congress

            Elected and non-elected officials have long understood how difficult it is to control the actions of administrative agencies.  Grants of discretion, uncertainty, divergent values, and information asymmetry that favor bureaucrats make control problematic.  For their part, elected officials have devised a number of mechanisms to overcome these and other obstacles to control.  Traditionally, Congress and the president have relied on budget authority, oversight hearings and appointment powers in an effort to ensure that agencies implement laws as intended.  The record to date, however, suggests traditional means of control are limited in their ability to detect and sanction shirking on the part of agency officials (McCubbins and Schwartz, 1984; McCubbins, Noll, and Weingast, 1985; Meier, 1999).        

In recent years, scholars have learned that political actors turn to additional mechanisms of control in order to mitigate the problems of information asymmetries and agency shirking.  McCubbins and Schwartz (1984), for example, argue that members of Congress engage in fire alarm oversight to reduce and shift the costs of monitoring agencies.  Similarly, McCubbins, Noll, and Weingast (1987) contend that elected officials use administrative procedures to enfranchise key constituents and structure agency decision-making.

Scholars have also discovered that political actors vary in terms of their preferences for particular control mechanisms.  Moe (1989) argues, for example, that where a single interest group or a coalition of groups dominate congressional policy making, those interests will attempt to put into place personnel, rules, and organizational structures that protect group interests in the future by “stacking the deck” in the groups’ favor and by insulating the agency from political influence.  In contrast, groups that were not part of the original winning coalition will seek organizational rules and structures that make subsequent agency decisions more transparent and more amenable to political interference.  Interestingly enough, Moe contends that members of Congress have few a priori structural preferences, choosing instead to adopt rules and procedures that meet the needs of dominant interests.  Presidents do, however, prefer rules that place administrative activities within the executive branch (versus independent regulatory commissions), allow Presidents to appoint key agency personnel, or require central clearance and coordination through the Office of Management and Budget (OMB).

These lessons were apparently not lost on House Republicans in the 104th Congress as they made regulatory reform a key element of the Contract with America.   In seeking to roll back the rate of regulation, House leaders relied on a mix of control mechanisms, including traditional mechanisms like budget cuts and a variety of deck stacking strategies (see Table 1).  What is particularly intriguing is how House Republicans often mixed the two.  In the process of paring agency budgets, for example, House leaders included dozens of riders aimed at limiting the reach of regulatory agencies.  In its first swipe at EPA’s budget, for example, Republicans not only sought a 32 percent funding cut, but offered 17 riders that would have limited EPA’s regulatory scope (Katz, 1995).  In a similar fashion, OSHA appropriations were initially cut by 16 percent and language was inserted that would have limited the agency’s reach as well (Hager, 1995).

Table 1 here

Appropriation politics aside, the main thrust of the Republican assault was a series of attempts to stack the deck in favor of those business and economic interests who bear the costs of federal regulation.  More precisely, since regulatory structures and rules were already in place in 1994, Republican leaders sought to “restack” the deck by a) enfranchising business interests while disenfranchising others and b) putting into place rules and structures that would make it more difficult for regulatory agencies to act.   In that latter regard, House Republicans quickly passed the Regulatory Transition Act in February of 1995.   In an effort to provide immediate relief to small and large businesses, the bill placed a 13-month moratorium on the implementation of regulations issued after the November, 1994 elections.  Other attempts to slow down the pace (and costs) of regulation were somewhat subtler. The reauthorization of the Paperwork Reduction Act, for example, aimed at minimizing the costs of complying with federal regulations by reducing federal reporting requirements and establishing the Office of Information and Regulatory Affairs (OIRA) in OMB to oversee those reductions. 

More typically, reform efforts sought to slow the pace of federal regulation by requiring federal agencies to conduct more thorough (and time-consuming) risk and cost-benefit analysis and to take that analysis into account in drafting new rules. The Unfunded Mandate Reform Act (P.L. 104-4), for instance, required agencies to prepare cost-benefit analysis of any regulation with estimated costs of $100 million or more and to gauge the regulation’s impact on the economy.  The centerpiece of Republican efforts to force agencies to attend to the economic impacts of regulation, however, was the Job Creation and Wage Enhancement Act (H.R. 9).  Under the proposed legislation, agencies were required to conduct detailed risk-assessment and cost-benefit analyses of new regulations and required OMB to conduct additional comparative risk analysis.  The proposed legislation would also have revised the Administrative Procedures Act so that a “major” rule would now be any rule that imposed private sector costs in excess of $50 (versus $100) million. 

House leaders also sought to provide regulatory relief by enfranchising key groups (and occasionally disenfranchising others) and by making rule-making more permeable and subject to congressional, judicial, and private review.  A good example of this was legislation introduced in the House and Senate, the Occupational Safety and Health Reform and Reinvention Act, that would have allowed firms to create their own workplace safety guidelines and hire outside inspectors to enforce those rules.  The House and Senate bills would have also have repealed the requirement that OSHA inspect a workplace whenever an employee filed a complaint. 

Along the same lines, the Small Business Regulatory Enforcement Fairness Act (P.L. 104-121) sought to provide regulatory relief by increasing the opportunities for the courts, Congress, industry representatives and other agencies to review agency rulemaking on procedural grounds.  The act requires, for example, that new rules be submitted to the Congress and the General Accounting Office (GAO) before the rule goes into effect.  The GAO must then prepare a report ensuring that the agency has complied with various procedural requirements (e.g. cost-benefit or risk analysis or the Unfunded Mandates Act).  Congress has 60 days to block the regulation by passing a joint resolution and sending it to the president for his signature.  If the president signs off, the regulation is nullified.  The act also makes it easier for small businesses to seek judicial relief where firms feel agency officials have not complied with the provisions of the Regulatory Flexibility Act.  Under the act, federal courts can order agency compliance and halt agency enforcement of the challenged regulation.  In addition, the Small Business Fairness Act also requires regulatory agencies to submit proposed rules to the Small Business Administration (Garcia, 2000).  Finally, the act directs the Small Business Administration to create within each region a small business ombudsman and “Fairness” board responsible for monitoring agency excesses (Eisner, Worsham, and Ringquist, 2000).  The immediate effect of all these provisions, of course, is to slow the rate at which agencies are able to issue new regulations by affording small businesses multiple opportunities to challenge regulations on procedural grounds.

            In the final analysis, much of what was proposed by the Republican leadership fell victim to a variety of forces including House-Senate differences, the budget standoff with President Clinton, and the ability of environmental and other groups to mobilize public opposition.  Nonetheless, what did pass provides firms some relief from regulation, expands the opportunity for regulatory review, and requires regulatory agencies to attend more carefully to the costs and benefits of their regulations. Moreover, in the process of drafting the legislation and moving it through the House and Senate, congressional Republicans set a tone that agencies could hardly ignore, at least in the short run.

            Our earlier analysis (Hedge and Johnson, 2002) of regulatory change at four federal agencies – the Environmental Protection Agency (EPA), the Office of Surface Mining (OSM), the Equal Employment Opportunity Commission (EEOC), and the NRC – indicates that Republican efforts had an immediate, if short-lived effect.  In every instance there was a noticeable decline in regulatory vigor in the wake of the 1994 elections that lasted just one or two years.  In accounting for those trends, we speculated that the timing and nature of our agencies’ response to congressional pressures was a function of bureaucratic dynamics and the larger drama of national politics. 

            One key question was why the four agencies responded at all to congressional discontent in the absence of any legal reason to do so.  Much of what the Republicans offered in the way of regulatory reform failed to pass and what did pass was signed into law months after the four agencies had begun reversing the rate of federal regulation.  Why didn’t Clinton appointees simply “wait out” the Republican challenges in the hopes that Senate moderates or the President would block those efforts (which is exactly what occurred)?  We suspect that part of the answer lies within the agencies themselves. 

Recent analysis by Eisner and Meier (1990), Brehm and Gates (1997), Golden (2000) and others remind us that bureaucratic responsiveness is as much about the predispositions of agents as it is the resources and efforts of their principals. While we can only speculate about what motivated political appointees and career civil servants at each of our agencies in the wake of the Republican revolution, we suspect that their responses reflected the kinds of considerations that Golden and others observed during the Reagan years -- individual and agency self interest mixed with professional norms required our agencies to respond to a changing political climate.  Sensing that political control had shifted from the White House to the Congress and given the apparent electoral mandate Republicans had received, senior agency officials no doubt concluded that it was in their agency’s self interest to adapt to the new political realities.  As for career civil servants, many of whom had served in the Reagan-Bush administrations, responding to (indeed anticipating) the new Republican majority was likely driven by “…their position in the political system, their sense of duty, and their code of conduct (Golden, p. 155).”

We also surmised that the timing and nature of the four agencies’ response to congressional pressures had much to do with larger presidential politics.  The 1994 elections were not only devastating for congressional Democrats, but for Bill Clinton as well.  Most analysts agree that the midterm elections were as much a rebuke to the Clinton presidency as they were to the Democratic majority.  What that meant for regulatory agencies during the first year or so of the Republican revolution is that they had to operate without any cover or protection from a White House that was struggling with its own political legitimacy.  Only when Bill Clinton was able to reestablish some control over the national political agenda did federal agencies feel capable of resisting congressional pressures.

In a broader sense, the evidence suggested to us that congressional-bureaucratic relations are conditioned by the larger drama of national politics, including the ongoing struggle between Congress and the president. Where presidents and Congress are at odds, agencies may feel less able to resist congressional pressures if there is a sense that the administration is politically vulnerable.  In addition, when presidents are “distracted” by their own problems (e.g. Clinton’s impeachment) agency officials may find themselves dealing with Congress on their own and without presidential support.  All of that led us to hypothesize that agencies will be more inclined to comply with congressional preferences when the president’s political and public standing was at its lowest, in this case where the president’s popularity was low and during the impeachment process in late 1998 and 1999.   

To test those hypotheses, we collected quarterly data on presidential popularity and regulatory outputs at OSHA and the NRC.  The presidential popularity data are the quarterly presidential approval ratings reported by the Gallup organization over the eight years of the Clinton Administration.  The enforcement data for OSHA and the NRC are from 1993-2000.  The OSHA data were collected using OSHA's Integrated Management Information System (IMIS).  The source of the information in IMIS is the local federal or state office in the geographical area where the activity occurred.  We collected data for every OSHA office in the United States.  The measure we use to estimate OSHA regulatory activity is the number of inspections performed by OSHA.  The data were collected by Standard Industrial Classification (SIC) Code.  For this analysis, we focus on OSHA inspections for Agriculture, Forestry, and Fishing (SIC Code 1), Mining (SIC Code 2), Construction (SIC Code 3), and Manufacturing (SIC Code 4).  These four units comprise approximately 73% of all OSHA inspections performed. 

The NRC data for this analysis were collected from the Office of Enforcement's Quarterly Progress Reports, Semi-Annual Reports, and Annual Reports.  The measures we use to estimate NRC regulatory activity include the number of Notices of Violation (NOV), number of Orders, number of Civil Penalties, and the dollar amount of civil penalties.  The NRC issues all these actions to four different entities, nuclear reactor licensees, medical licensees, materials licensees, and individuals working in any of these fields.  The NRC issues notices of violation when an infraction in the law is not particularly severe and does not warrant a financial penalty.  Civil penalties are assessed when infractions in the law are determined to be severe.  Orders usually, but not always, are given by the NRC when they are revoking or suspending an operating license to either an individual or a company. 

Findings

            Figures 1-6 provide an initial glimpse into how the agencies responded to the often-dramatic political events of the 1990s.  In every instance, there is a sharp decline in enforcement levels following the “Republican Revolution.” That decline in regulatory vigor typically was short-lived, however, and was often followed by an equally dramatic increase in the rate of enforcement.  Those tendencies are especially pronounced in the case of the NRC (see Figures 1-3).  Indeed, the dollar value of civil penalties and the number of notices of violation issued by the NRC in 1996 and 1997 actually exceeded those levels in the pre-revolution era. 

Figures 1-6 here

The data also indicate that the decline in federal regulation actually began prior to the midterm elections of 1994.   We suspect some of that decline is simply the agencies’ anticipation of the Republicans’ midterm victories. But there is also reason to believe that the agencies were responding to the president’s own political problems.  As Figure 7 indicates, the president’s public approval hit its lowest levels in the summer of 1994.

Figure 7 here

The data in the figures also indicate that in some instances the level of regulation also fell sharply later in the decade.  In Figures 1 and 2 the level of civil penalties at the NRC dropped sharply in 1998 and continued to decline through the end of the Clinton administration.  In a parallel fashion, the number of OSHA manufacturing inspections began to decline in 1997 and generally moved downward through 2000 (see Figure 6).  How do we account for those declines?  In the case of the NRC and possibly OSHA, it could be the case that regulatory officials were reacting to the 1998 midterm elections and/or the House Impeachment vote later that year.  It is possible, for instance, that White House and agency officials may have decided to regulate less vigorously going into the 1998 midterm elections, particularly given the beating they had taken just 4 years earlier.  It is also possible agency officials sensed that the President was heading into another period of vulnerability as the Monica Lewinsky scandal and the impeachment process began heating up. 

Tables 2-5 analyze the effects of presidential standing, agency appointments, and the Republican Revolution on bureaucratic outputs by analyzing OSHA enforcement data for four distinct inspection categories: agriculture, mining, construction, and manufacturing.  The model specified is a Bayesian hierarchical AR(1) model, meaning that we specify Bayesian priors for unknown parameters, there is a random effects assumption, and the time series of enforcements is assumed to have a one-period autoregressive dependency. 

The Bayesian process of data analysis is characterized by three primary attributes: a willingness to assign prior distributions to unknown parameters, the use of Bayes rule to obtain a posterior distribution for unknown parameters on observable data, and the description of inferences in probabilistic terms (Gill 2002).  The primary philosophical foundation of Bayesian inference is the consideration of both observables and parameters as random quantities.  One advantage of this approach is that there is no restriction to building complex models with multiple levels and many unknown parameters, as we do here.  The posterior distribution of the coefficients of interest is obtained by multiplying the prior by the likelihood function and normalizing.  In a Bayesian hierarchical model, such priors also have priors (hyperpriors), so there are additional terms in the product but no real increase in complexity.  When these calculations are not straightforward, then the full engine of Markov chain Monte Carlo estimation is employed (Liu 2001, Chen, Shao, and Ibrahim 2000).

Since enforcements are recorded as counts, the lowest level of the model hierarchy is a Poisson specification for the outcome variable, Yt where t indexes the 32 time periods.  We further stipulate that the linear additive component of the model have a two-tier error term, reflecting the AR(1) assumption.  Thus the basic model is described in “stacked” notation as:

            Yt ~ Poisson (μt)

            μt = Xt β + εt

            εt = ρ ε{t-1} + νt.

This means that that the error term in time t is functionally dependent on the error term at time t-1, as well as having its own zero-mean component, νt.  The matrix X contains t=32 rows of explanatory variables along with a leading column of 1's.  These explanatory variables are: a two-digit approval rating for Clinton, a date index, a dichotomous indicator for the fourth quarter of 1994, a dichotomous indicator for the first quarter of 1995, and a dichotomous variable for the first quarter of 1998 and the first quarter of 1999.  Here β is a k=7 length vector of coefficient estimates.

Since the model is Bayesian we stipulate the following prior structure:

            ρ ~ Uniform(0,1)

            νt ~ Normal(0,10)

            bk ~ N(0,10)  for k=1,...,7.

 

This represents a relatively uninformed prior structure where little substantive information is interjected into the posterior from these prior distributions.  However, due to the hierarchy of the prior specifications, analytical calculation of the posterior is not straightforward and we implement a Gibbs simpler (Gelfand and Smith 1990) with the user-friendly package winbugs.

            The Markov Chain is run for 10,000 iterations, after a burn-in period of 10,000 iterations, recording parameter values as the chain length accumulates.  The burn in period is sufficiently conservative, and is started at several points.  The primary concern with MCMC results is assessment of convergence to the desired limiting distribution, and we ran several of the standard diagnostics with no apparent problems evident.

The results are given in the following form with the posterior mean and standard deviation from the Markov chain simulations.  These can be interpreted as standard logit results, although a full Bayesian treatment would describe the posterior in greater detail with quantiles.

Tables 2-5 here

The most striking finding in the tables of course is the consistent impact of presidential approval on the rate of inspections at OSHA.  In each area of OSHA enforcement, increases in the President’s approval ratings produce higher inspection levels.  In contrast, the Republican Revolution intervention variables had no reliable impact. While it is clear that the rate of regulation declined following the Republican takeover in 1994 the evidence would suggest that those declines had as much to do with the President’s overall standing as they did with the change in the partisan control of Congress.  From the agency’s perspective, it must have seemed as if their “boss” had few friends in Washington and among the public.  That had to have a chilling effect on regulatory vigor. There is also little support in the data for the notion that any decline in the rate of enforcement at OSHA was an agency response to the impeachment process.   In only one instance, construction inspections, was impeachment associated with lower levels of enforcement. [i]

Conclusions

            The Republican Revolution in the mid-nineties provides an excellent vantage for viewing political control of bureaucracy.  As the nation moved through one of its most politically tumultuous periods, federal agencies faced a Congress committed to drastically recasting the character of regulation and a President that was struggling to regain the legitimacy of his office.  Not surprisingly, our agencies responded to those pressures and cross-pressures.  As our data make clear, the rate of federal regulation during the Clinton administration responded to both congressional pressures and the President’s political standing.

What is so striking about our findings is not the fact that two federal agencies responded to political influences (that’s old news) but how and when they went about that.  First, despite the conventional wisdom that says bureaucracies move at glacier speed, each of the agencies proved remarkably capable of quickly reversing the rate of federal regulation, not just once but in many instances 2-3 times in a relatively short period of time.  Second, our findings suggest that political control, at least in the short term, is not just about finding the right tools of control. Scholars and policy makers have invested a fair amount of effort in discovering those mechanisms that are most effective in achieving political control (e.g. deck-stacking, budgets, appointments).  There is no shortage of evidence that the tools of political control matter.  At the same time, the evidence from the early days of the Republican revolution suggest that political control can be achieved even in the absence of any legislative mandate to do so.  In the short run at least, the challenge for Congress is not to find the “right” control mechanisms but to send a clear and unambiguous signal of its dissatisfaction with agency behavior and a willingness to act on those concerns.  No doubt, the quickness with which House Republicans passed sweeping regulatory reforms in the first few weeks of the 104th Congress offered such an unequivocal message to agency officials. 

Our findings also underscore the need to consider the influence of multiple principals and how the actions of each condition the ability of others to achieve political control. Political control is not an additive process in a political system in which powers are shared between branches of government.  Where the preferences of political principals differ, as they did most dramatically in the mid-nineties, an increase in political control for one set of principals necessarily comes at the expense of the ability of others to influence agency behavior.  For the Clinton White House that became very clear in the wake of the 1994-midterm elections.  At the same time, our findings provide some support for the notion that the Congress’ ability to “move” federal agencies during the Clinton administration was conditioned in large part by the President’s standing with the public.  When the President’s approval ratings were low, OSHA and the NRC were much more inclined to respond to congressional demands for less regulation. Once the president’s popularity resurged, the two agencies were much less compliant.



 

 


           


 

 


 

 


 

 


 



 

 

 


 





 



 


Table 1 -- The Republican Regulatory Agenda -  104th Congress

 

Regulatory Transition Act of 1995

Purpose:  place a 13-month moratorium on the implementation of regulations issued after November 20, 1994

Outcome:  The bill was overwhelmingly passed in the House on February 24, 1995.  However, it was later defeated in the Senate on June 16, 1995.

 

Unfunded Mandates Reform Act of 1995 (P. 104-4)

Purpose:  to prohibit Congress from imposing mandates on private-sector actors and state and local governments without providing funding for their implementation

Outcome: Senator Dirk Kempthorne (R-ID) introduced the bill on January 4, 1995 and President Clinton signed the bill into law on March 22, 1995

 

The Job Creation and Wage Enhancement Act of 1995 (H.R. 9)

Comprehensive Regulatory Reform Act of 1995 (S. 343)

Purpose:  regulatory overhaul in the form of risk-assessment and cost-benefit analysis procedures, regulatory relief, and property rights

Outcome:  H.R. 9 passed in the House on March 3, 1995, but stalled in the Senate. 

 

Paperwork Reduction Act (PL 104-13)

Purpose:  Reauthorization of Paperwork Reduction Act to reduce the number of hours required to comply with agency reporting requirements

Outcome:  Signed into law in May, 1995.

 

Revamping OSHA (S. 1423, H.R. 3234)

Purpose:  Allow employers to create their own workplace safety rules and hire outside inspectors to enforce them; allow Secretary of Labor to waive fines for small firms who correct violations quickly; repeal the requirement that OSHA inspect every workplace where an employee had filed a complaint

Outcome:  Legislation failed to reach the floor of either chamber by the end of the session.

 

Small Business Regulatory Fairness Act (P.L. 104-121)

Purpose:  Ease the regulatory costs on small firms.  Created Small Business Regulatory Fairness Boards; provides judicial review/relief of agency compliance with flexibility requirements; requires Congressional review of agency cost-benefit and flexibility analysis. 

 

Appropriation Politics

Purpose:  Dozens of appropriation riders attached to appropriation bills aimed at slowing or halting enforcement of existing environmental regulations; operating and enforcement budgets cut across several regulatory agencies

Outcome: Riders stripped from appropriations bill and Congress agrees to reinstate many of its earlier cuts following the budget standoff with President Clinton.

 

 

 

Table 2:  OSHA Manufacturing Inspections

 

 

Mean        Standard Deviation

 

              

Constant                                                           -0.11380176              0.1333407308

 

Clinton’s Approval Rating                               0.00427709               0.0009393287

 

Date(time) Index                                              0.01572576               0.0096669655

 

Appointment                                                     1.54856162               1.5529008888

 

1994 Election (4th quarter 94)                         0.69182974               0.5267894977

 

New Republican Congress (1st quarter 95)   1.01732339               0.3938731411

 

Clinton Impeachment                                       0.28701456               0.4154827336

 

 ρ                                                                        0.38876220               0.0006072209

 

 

 

 

 

 

 

Note:  The dependent variable is the number of OSHA manufacturing inspections.  The results are given in the following form with the posterior mean and standard deviation from the Markov chain simulations.  These can be interpreted as standard logit results, although a full Bayesian treatment would describe the posterior in greater detail with quantiles.

 


 

 

Table 3:  OSHA Construction Inspections

 

 

                                                                                    Mean        Standard Deviation

        

Constant                                                          -0.134311762             0.0775517482

 

Clinton’s Approval Rating                              0.010571768              0.0019916266

 

Date(time) Index                                            -0.003533841             0.0025043148

 

Appointment                                                    0.590685917              0.4051884727

 

1994 Election (4th quarter 94)                       -0.357096757             0.5110659629

 

New Republican Congress (1st quarter 95)   0.030682691              0.3291544513

 

Clinton Impeachment                                     -1.696949687             0.2701719367

 

ρ                                                                        0.385892845              0.0004437062

 

 

 

 

 

Note:  The dependent variable is the number of OSHA inspections.  The results are given in the following form with the posterior mean and standard deviation from the Markov chain simulations.  These can be interpreted as standard logit results, although a full Bayesian treatment would describe the posterior in greater detail with quantiles.

 

 


 

 

 

 

 

Table 4:  OSHA Mining Inspections

 

 

                                                                                    Mean      Standard Deviation

              

Constant                                                           4.53400872                 0.71994202

 

Clinton’s Approval Rating                              -0.02002073                0.01599115

 

Date(time) Index                                              0.15350681                 0.03523012

 

Appointment                                                    -1.13325142                0.88865586

 

1994 Election (4th quarter 94)                        -0.32931913                1.08661420

 

New Republican Congress (1st quarter 95)   -1.15965484                1.75373611

 

Clinton Impeachment                                      -3.05787029                1.60138587

 

ρ                                                                         0.87987910                 0.15481072

 

 

 

 

Note:  The dependent variable is the number of OSHA inspections.  The results are given in the following form with the posterior mean and standard deviation from the Markov chain simulations.  These can be interpreted as standard logit results, although a full Bayesian treatment would describe the posterior in greater detail with quantiles.

 

 

 


 

 

 

Table 5:  OSHA Agriculture, Forestry, and Fishing Inspections

 

 

                                                                                    Mean       Standard Deviation

              

Constant                                                           -0.30127509               0.505143677

 

Clinton’s Approval Rating                               0.04519125                0.009834378

 

Date(time) Index                                              0.08697005                0.022693548

 

Appointment                                                     0.24902888                1.848437952

 

1994 Election (4th quarter 94)                         0.46240081                2.187819571

 

New Republican Congress (1st quarter 95)   2.59225597                2.958968838

 

Clinton Impeachment                                       2.11548228                1.322426063

 

ρ                                                                         0.45389673                0.041672199

 

 

 

 

Note:  The dependent variable is the number of OSHA inspections.  The results are given in the following form with the posterior mean and standard deviation from the Markov chain simulations.  These can be interpreted as standard logit results, although a full Bayesian treatment would describe the posterior in greater detail with quantiles.

 

 

 


Bibliography

 

Banks, Jeffrey S. and Barry R. Weingast.  1992.  “The political control of bureaucracies under asymmetric information”.  American Journal of Political Science; 36:2(509-525).

 

Benenson, Bob. “GOP Sets the 104th Congress on New Regulatory Course,” June 17, 1995. CQ Weekly Report. pp. 1693-1701.

 

Brehm, John and Scott Gates. 1997.  Working, Shirking, and Sabotage: Bureaucratic Response to a Democratic Public.  Ann Arbor:  University of Michigan Press.

 

Briggs, Michael.  April 29, 1992.  “EEOC Declining Under Bush, Senators Told.”  Chicago Sun-Times, News Section, page 9.

 

Brownstein, Ronald.  May 22, 1994.  “Key Civil Rights Post Left Empty As Search Falters.”  The New York Times, Part A, Page 1. 

 

Dodd, Lawrence and Richard Schott. 1979. Congress and the Administrative State.  New York: John Wiley, 1979.

 

Eisner, Marc Allen.  1993.  Regulatory Politics in Transition.  Baltimore, MD:  The Johns Hopkins University Press.

 

Eisner, Marc Allen and Kenneth Meier. 1990 “Presidential Control and Bureaucratic Power:  Explaining the Reagan Revolution in Antitrust.”  American Journal of Political Science.  34:1 (269-287).

 

Eisner, Marc Allen, Jeff Worsham, and Evan J. Ringquist. 2000. Contemporary Regulatory Policy.  Boulder, CO:  Rienner Publishers Inc.

 

Marissa Golden. 2000. What Motivates Bureaucrats?  Politics and Administration During the ReaganYears.  New York: Columbia University Press, 2000.

 

Gormley, William T. Jr.  1989.  Taming the Bureaucracy:  Muscles, Prayers, and Other Strategies.  Princeton, NJ:  Princeton University Press.

 

Grimsley, Kirstin Downey.  February 11, 1996.  “EEOC Chief Voices Frustration Over Case Backlog, Budget Cuts.”  The Washington Post, Section A, Page A04.

 

Hansen, Wendy L., Johnson, Renée J., and Isaac Unah.  1995.  “Specialized Courts, Bureaucratic Agencies, and the Politics of U.S. Trade Policy”.  American Journal of Political Science; 39:3(529-558).

 

Harris, Richard A. and Sidney M. Milkis.  1996.  The Politics of Regulatory Change:  A Tale of Two Agencies, 2nd Ed.  New York, NY:  Oxford University Press.

 

Hedge, David and Renee Johnson. 2002. “The Plot that Failed:  The Republican Revolution and Political Control of the Bureaucracy. “ Journal of Public Administration Research and Theory. (July, 2002)

 

Hedge, David and Michael J. Scicchitano.  1994.  “Regulating in Space and Time:  The Case of Regulatory Federalism.”  The Journal of Politics, 56(1):134-153.

 

Hook, Janet. March 26, 1995.  “Weary House Republicans Set Breakneck Legislative Pace.”  The Los Angeles Times:  Part A; Page 1.

 

Garcia, Regelio.  May 11, 2000 “Federal Regulatory Reform:  An Overview.”  Congressional Research Service Issue Brief.  Washington D.C.: The National Council for Science and the Environment.

 

Golden, Marissa Martino.  2000.  What Motivates Bureaucrats?:  Politics and Administration During the Reagan Years.  New York, NY:  Columbia University Press.

 

Katz, Jeffrey. August 5, 1995. “House Revives EPA Restrictions Before Passing VA-HUD Bill.”  CQ Weekly Report.  Pp.2366-2369

 

Krause, George A.  1999.  A Two-Way Street:  The Institutional Dynamics of the Modern Administrative State.  Pittsburgh, PA:  University of Pittsburgh Press.

 

Labaton, Stephen.  October 31, 1993.  “Administration Leaves Top Civil Rights Jobs Vacant.”  The New York Times, Section 1; Page 18.

 

Lane, Jan-Erik.  1995.  The Public Sector:  Concepts, Models, and Approaches, 2nd Ed.  London, UK:  Sage Publications, Ltd.

 

Mathews, Jessica.  November 27, 1995.  “Battle for the Environment.”  The Washington Post:  Page A21.

 

McCubbins, Mathew D. and Thomas Schwartz.  1984.  “Congressional Oversight Overlooked:  Police Patrols versus Fire Alarms.”  American Journal of Political Science, 28:1(165-179).

 

McCubbins, M., Noll, R., and B. Weingast.  1989.  “Administrative Procedures as Instruments of Political Control.”  Journal of Law, Economics, and Organization, 3:2(243-277).

 

Moe, Terry M.  1984.  “The New Economics of Organization.”  American Journal of Political Science, 28:4(739-777).

 

Moe, Terry M.  1985.  “Control and Conflict in Economic Regulation:  The Case of the NLRB.” American Political Science Review.  79:1094-1117.

 

Scholz, John T., and Feng Heng Wei.  1986.  “Regulatory Enforcement in a Federalist System.  American Political Science Review, 80:4(1249-1270).

 

Spotts, Peter N.  June 25, 1996.  “By-the-Book NRC Chief Starts to Shake Up Old-Boy Agency.”  Part Four of Series: The NRC -- Watchdog of Pager Tiger? The Christian Science Monitor, Page 1.   

 

Wald, Matthew L.  June 23, 1992.  “Nuclear Agency’s Chief Praises Watchdog Groups.”  The New York Times, Section A; Page 13.

 

Waterman, Richard W., Rouse, Amelia, and Robert Wright.  1998.  “The Venues of Influence: A New Theory of Political Control of the Bureaucracy”.  Journal of Public Administration Research and Theory; 8:1(13-38).

 

Wood, B. Dan.  1988.  “Principals, Bureaucrats, and Responsiveness in Clean Air Enforcements.”  American Political Science Review, 82:1(213-234).

 

Wood, B. Dan and Richard W. Waterman.  1991. “The Dynamics of Political Control of the Bureaucracy.” American Political Science Review.  85: 801-828.

 

Wood, B. Dan and Richard W. Waterman.  1994.  Bureaucratic Dynamics:  The Role of Bureaucracy in a Democracy.  Boulder, CO:  Westview Press.

 



[i]  We are still awaiting data from the NRC in order to complete our time series and perform a similar analysis with those data.