FAS Homepage | Drug Policy | Index | Search | Join FAS
FAS Drug Policy Project Bulletin

FAS Drug Policy Analysis Bulletin

Issue Number Ten
February 2004


Fuzzy Math: Why the White House Drug Control Budget Doesn't Add Up

John M. Walsh
Senior Associate for the Andes and Drug Policy, Washington Office on Latin America

Note: Subscriptions to the FAS Drug Policy Analysis Bulletin may be requested by sending email to fas@fas.org, please be sure to reference the publication title

The federal drug control budget prepared by the Office of National Drug Control Policy (ONDCP), a staple of the drug policy debate, has been nothing if not predictable: year after year, the budget has increased, and year after year the ratio between "demand reduction" and "supply reduction" spending has remained roughly one-third to two-thirds. The basic theme has been that for every new dollar spent on demand reduction, two new dollars would be spent to curb supply. President Reagan's final drug budget (fiscal year 1989) totaled $6.5 billion, and allocated 30% to demand reduction vs. 70% to supply reduction. The budget grew to $11.9 billion by the end of George H. W. Bush's presidency (FY1993), with a 33% to 67% demand-supply allocation. Under President Clinton, the budget reached $18.1 billion (FY2001), with the same 33% vs. 67% allocation. Domestic law enforcement has always been the single largest component of the budget; in FY1994, it surpassed 50% of the budget for the first time, and has accounted for most of the budget ever since.

Now, with a "restructured" budget request for FY2004, the Bush administration appears to have reined in federal spending on the "war on drugs" and come close to a 50-50 split between demand and supply reduction. The Administration's $11.7 billion FY2004 request would slash $7.5 billion from the $19.2 billion FY2003 request, with nearly 90% of the reductions coming out of supply control functions (domestic law enforcement, interdiction and international programs). The proportion of the total $11.7 billion budget devoted to demand reduction (treatment and prevention) would approach 47%, after decades of being stuck at one-third.1 When previewing its planned budget restructuring in February 2002, ONDCP noted that the new method "might" lower by "several billion dollars" the amount of funding reported in the drug control budget,2 but ONDCP offered no forecast as to the likely demand/supply breakdown of the restructured budget.

What to make of it? Should critics of the emphasis on arrests and incarceration and proponents of greater investments in prevention and treatment be cheered by the Administration's new budget? To the contrary. President Bush's ONDCP "restructured" its drug control budget with the laudable goal of providing "a more realistic basis for policymakers to consider tradeoffs between spending for prevention, treatment and law enforcement programs."3 But despite some improvements, ONDCP has achieved the opposite of its expressed intent. The restructured budget is a notably less realistic portrayal of federal drug control spending than its predecessor, and is less useful than ever as a basis for policy decisions.

The key to making sense of how ONDCP restructured the budget--and why the new version is an even worse depiction of federal drug control spending than previous versions--is to first understand the artificial nature of the drug control budget itself. ONDCP has little authority, formally or de facto, over actual drug control spending by federal departments and agencies. By statute, ONDCP must approve the drug budgets of agencies, but historically it has had only slight influence on the final figures. The budget presented by ONDCP each year is not prescriptive, but merely describes what is being spent by various agencies, based on various calculations estimating what proportion of a given agency's spending is drug-related (spending for some agencies, such as the DEA and NIDA, are considered 100% drug-related). As such, the budget is subject to two basic kinds of errors: wrongly including spending that should not appear in the budget, and wrongly excluding spending that should appear.

On the one hand, showing an agency's spending in the budget when it should not be counted results in overstating total drug control spending and exaggerating the level of spending on particular functions (e.g., treatment, prevention, etc.). For example, a 2000 RAND study--commissioned by President Clinton's ONDCP--found that the methods used by the Veterans Administration (VA) and other agencies to calculate treatment spending tended to result in substantial overestimates.4 Based on RAND's findings, readjusted ONDCP budget figures revealed that treatment spending had been overstated by an average of nearly $410 million per year from FY1991-FY2000, accounting for one-seventh of total ONDCP-estimated treatment expenditures during this period.

On the other hand, ONDCP's budget can present an incomplete picture of federal drug control spending when categories of spending that should be shown are left out. Such exclusions leave policymakers and the public with a misleadingly partial account of what is being spent on drug control. It bears emphasis that, given the artificial nature of the federal drug control budget, just because a category of spending is left out of the budget presented by ONDCP does not mean that the spending does not occur. Such errors of exclusion, on a very large scale, characterize ONDCP's current restructuring effort.

Some attempt to reform and streamline the federal drug control budget was very much in order. As illustrated by the years of exaggerated treatment spending estimates, there is much room for improving the methods used to estimate how much of a given agency's spending should be considered related to drug control and presented in ONDCP's budget. Moreover, the expressed philosophy guiding ONDCP's restructuring effort makes good sense: The budget should be limited to "those agencies or accounts that have been, or should be, the principal focus of drug control policy."5 Historically, numerous agencies with only the most peripheral involvement in drug control have been included in the budget. No one will miss the meager line items appearing in previous drug budgets for the U.S. Forest Service, the Bureau of Land Management, the National Park Service, etc. However, no longer reflecting these agencies in the budget makes little difference in overall spending, or in its demand vs. supply allocation. For example, nine agencies that do not appear at all in ONDCP's restructured FY2004 budget request accounted for only $39.4 million in spending in the Administration's FY2003 request. Removal of these items comprises only ½ of 1% of the overall $7.5 billion reduction between FY2003 and FY2004.

The big differences come with respect to spending on domestic law enforcement, the single most dominant function within the budget. Here, the logic behind ONDCP's decisions about what stays in the budget, and--more to the point--what gets left out, is extremely dubious. The Administration's tortured logic, and its impact on the budget, is exemplified by its handling of spending by the Bureau of Prisons (BOP), the federal judiciary, and several Justice Department agencies involved in the prosecution, detention and incarceration of federal drug offenders (Federal Prisoner Detention, Immigration and Naturalization Service, U.S. Attorneys, U.S. Marshals, Criminal Division).

In the President's FY2003 budget request, drug control spending by these agencies amounted to nearly $4.4 billion. By contrast, the restructured FY2004 request reflects only $45 million in spending for the same agencies (for drug abuse treatment for BOP inmates). Nearly 90% of the $7.5 billion decrease from FY2003 to FY2004 is due to the removal of spending from supply control categories. Most of the decrease (58%) stems from the disappearance from the budget of items related to the prosecution and incarceration of drug offenders. Removing the Bureau of Prisons, other Justice agencies, and the federal judiciary goes a long way toward bringing what remains of the budget into closer balance with respect to demand vs. supply spending.

Justice Department Spending Reflected in ONDCP Budgets

 
(Budget Authority in Millions)
FY2003 Budget Summary FY2004 Budget Summary
FY2001 Final FY2002 Enacted FY2003 Request FY2002 Final FY2003 Request FY2004 Request
Assets Forfeiture Fund439.9360.0430.0***
U.S. Attorneys228.2244.6254.4***
Bureau of Prisons2,341.52,525.12,443.039.443.545.2
Community Policing374.7427.4653.3***
Criminal Division35.137.838.7***
Drug Enforcement Administration1,480.41,605.41,698.51,562.51,659.61,677.3
Federal Bureau of Investigation707.5415.5421.4***
Federal Prisoner Detention375.5429.4463.9***
Immigration and Naturalization Service525.0538.0713.4***
Interagency Crime and Drug Enforcement325.2338.6362.1446.5470.3541.8
INTERPOL0.30.30.3***
U.S. Marshals Service223.8255.1277.8***
Office of Justice Programs1,016.6962.6309.2893.2286.7301.5
Tax Division0.40.40.4***
Total8,074.18,140.18,066.52,941.52,460.12,565.8
* Not included in restructured budgets.

This does not mean that theses federal agencies will no longer be spending several billion dollars per year to incarcerate federal drug offenders; indeed, such spending can be expected to continue apace. It does mean that ONDCP will no longer present this spending as part of the annual federal drug budget. To understand the scale of spending on incarceration that ONDCP now considers irrelevant to its portrayal of federal drug control spending, consider that:

ONDCP acknowledges that removing incarceration-related spending from the budget represents a significant departure from the past, but defends its decision by arguing that, "[a]lthough these costs are real to society, they do not factor into the core of drug law enforcement decisions made by national policymakers."9 Whether or not policymakers actually do try to factor the costs of incarceration into their drug law enforcement decisions is an empirical question, and ONDCP offers no evidence to support its assertion that they do not. What is clear, however, is ONDCP's belief that the costs of incarceration should not be a factor in drug policy deliberations.

Since this seems to be a less than self-evident proposition, what is the reasoning behind it? According to ONDCP, the drug budget was restructured to reflect "only those expenditures aimed at reducing drug use, rather than those associated with the consequences of drug use."10 Within this dichotomous framework, ONDCP considers incarceration to be a consequence of drug use. In this view, the costs of incarceration should be estimated and reported--like the economic costs to society of drug-related hospital ER visits or the spread of drug-related HIV/AIDS--but do not properly belong in the federal drug budget.

But spending on the incarceration of federal drug offenders is part and parcel of federal efforts to control drug use through law enforcement-the predominant feature of the federal drug control strategy, and of American drug policy in general. It is true that incarceration is sometimes a consequence of drug use (or, more typically in the case of federal incarceration, a consequence of distributing drugs), just as overdose is sometimes a consequence of drug use. But the possibility of incarceration as a consequence of drug use is the result of specific policies and laws, as well as the ongoing funding of investigations, enforcement operations, prosecutions and corrections. Prosecution and incarceration cannot intelligibly be separated from the earlier stages of enforcement (investigations, arrests).

Enforcement can work through arrests and seizures alone, but the various rationales offered for stringent drug enforcement--including punishment for its own sake and the crime-control goals of incapacitation and deterrence--all ultimately derive from the threat of prison terms. This is perfectly clear in practice. When federal drug enforcement agencies such as the DEA pursue arrests, they do so with the objective (and expectation) of prosecution, conviction and ultimately imprisonment. For example, U.S. attorneys investigated nearly 38,000 suspects on drug charges in 2001. The same year, more than 28,000 defendants had drug cases terminating in U.S. district courts; 92% of the defendants were convicted. Of the 25,000 drug offenders convicted and sentenced in U.S. district courts in 2001, 92% were sentenced to prison.11

The efficacy and cost-effectiveness of incarceration as a means of reducing drug use is, of course, debatable. But there can be no doubt that incarceration has been--and remains--a large, expensive, and high-profile component of the federal government's efforts to reduce drug use. Moreover, incarceration levels are the result of policy decisions, not just by the Executive (which decides the number of cases brought and how they are charged), but by Congress itself. Congress can--and regularly does--make statutory changes designed to increase sentencing severity, such as the Illicit Drug Anti-Proliferation Act of 200312 (aimed at raves and club drugs) and the Ecstasy Anti-Proliferation Act of 2000.13 Such legislation leads to more people in prison and higher levels of expenditure by the Bureau of Prisons. ONDCP cannot sensibly present a budget whose domestic law enforcement component excludes spending associated with prosecution and incarceration.

Incarceration also differs from drug use consequences such as overdose and the spread of HIV/AIDS in that the latter tend to be diffuse, with costs borne by various levels of government and different sectors of society. By contrast, spending on the incarceration of federal drug offenders is an identifiable, discrete component of federal government spending. Indeed, among the ten federal agencies considered by RAND in its 2000 report for ONDCP, the Bureau of Prisons' drug-related spending estimation methods were considered among the few to "reflect quite accurately the resources being expended." RAND did express concern that, if anything, the Bureau's estimation methods "could actually lead to some understatement of the BOP drug budget" (italics in the original).14

Beginning with the FY2004 request, ONDCP budget reports will only present figures calculated by the new method, including revised versions of historical budgets (i.e., what previous budgets would have looked like had the methods being introduced with the FY2004 request been applied then). On one hand, this will provide for a consistent time-series, allowing for direct comparisons between past and future budgets produced with the same methods. But by failing to also display the old budget time-series, ONDCP obscures the magnitude and impact of the changes it has made. Readers without knowledge of the budget history will literally not know what they are missing. For example, the President's FY2003 request included $740 million in drug-related prosecutions and corrections funding for the federal judiciary. But effective with the FY2004 request, ONDCP budgets will make no mention of the federal judiciary, not even with regard to past budgets. This item--and many similar enforcement-related functions--will have disappeared from the budget without a trace, even though the federal judiciary will surely continue to process thousands of drug offense cases each year.

When improved methods allow for significantly more accurate budgeting (or other forms of data collection), then breaking with the past is warranted, even though comparisons with past budgets become problematic. In this case, though, ONDCP's restructured budget creates more problems than it solves. Indeed, the dubious logic behind certain decisions--especially to eliminate spending associated with the incarceration of federal drug offenders--suggests that the Bush administration's real agenda was to at last defuse criticism of the historically imbalanced federal drug control budget. By simply removing billions in spending on domestic law enforcement from the budget, the federal drug control effort suddenly appears to be nearly evenly balanced between demand reduction and supply reduction, without having to spend a penny more on treatment and prevention. Moreover, the arrest-and-incarcerate strategy can be expected to continue apace, driving incarceration-related spending ever higher, but without ever again troubling policymakers by appearing in the ONDCP budget.

Endnotes

  1. ONDCP, National Drug Control Strategy, FY2004 Budget Summary, February 2003.
  2. ONDCP, National Drug Control Strategy, FY2003 Budget Summary, February 2002.
  3. ONDCP, National Drug Control Strategy, FY2004 Budget Summary, February 2003.
  4. Murphy, P. et al., Improving Anti-Drug Budgeting. Santa Monica, CA: RAND, 2000.
  5. ONDCP, National Drug Control Strategy, FY2003 Budget Summary, February 2002.
  6. Bureau of Justice Statistics (BJS), Federal Criminal Case Processing, 2001, January 2003.
  7. "Federal Bureau of Prisons QUICK FACTS" available at http://www.bop.gov/fact0598.html
  8. The total expected cost of federal imprisonment for the 23,000 newly sentenced federal drug offenders in 2001 was calculated by the author based on an average of 5.5 years to be served in prison at $27,850 in operating expenditures per inmate per year. See BJS, Federal Criminal Case Processing, 2001, January 2003 (for number of newly sentenced federal drug offenders); Time Served in Prison by Federal Offenders, 1986-1997, June 1999 (for time to be served); and State Prison Expenditures, 1996, August 1999 (for federal prison operating expenditures per inmate per year, updated by the author for inflation using the Bureau of Labor Statistics' "CPI Inflation Calculator," available at http://www.bls.gov).
  9. ONDCP, National Drug Control Strategy, FY2003 Budget Summary, February 2002.
  10. ONDCP, National Drug Control Strategy, FY2004 Budget Summary, February 2003.
  11. BJS, Federal Criminal Case Processing, 2001, January 2003.
  12. Public Law 108-21, Title VI, Section 608. April 30, 2003.
  13. Public Law 106-310, Title XXXVI, Subtitle C. October 17, 2000.
  14. Murphy, P. et al., Improving Anti-Drug Budgeting. Santa Monica, CA: RAND, 2000.

Drug Policy News

Litigation

Raytheon Co. v. Hernandez
The Supreme Court reversed a Ninth Circuit ruling that Raytheon Co. had violated the Americans with Disabilities Act by refusing to rehire Hernandez, a rehabilitated former employee fired 2 years earlier for failing a drug test. The Court disagreed with the Ninth Circuit's consideration of the impact, rather than intent, of the Company's unwritten mandate against rehiring terminated employees, stating that "[the] petitioner's no-rehire policy is a quintessential legitimate, non-discriminatory reason for refusing to rehire an employee who was terminated for violating workplace conduct rules."
http://www.ncd.gov/newsroom/news/r04-442.html

Canadian court upholds marijuana law
The Canadian Supreme Court recently upheld the national marijuana law, reducing the pressure on Parliament to address proposed changes to the law. The case was brought by three men charged with possession of small amounts of marijuana, who argued that criminalization of marijuana infringed on their rights and freedoms and violated their right of fundamental justice. In its ruling, the Court determined that Parliament is authorized to change the law, and Prime Minister Paul Martin has announced plans to introduce a bill similar to one introduced by former Prime Minister Jean Chretien that would eliminate criminal penalties for possession of less than 15 grams of marijuana.
http://www.cnn.com/2003/WORLD/americas/12/23/canada.marijuana.ap/

Prenatal drug exposure
The Supreme Court has decided not to review South Carolina v. Regina McKnight-the murder conviction and thirteen-year prison term of a young woman whose child was stillborn in 1999. An autopsy revealed cocaine metabolites in the child's system, although no medical evidence presented at trial showed that the stillbirth was caused by cocaine.
http://www.courttv.com/news/2003/1006/stillbirth_ap.html

Legislation

Alaska 2004 ballot initiative
A proposal to legalize marijuana for private use has met the voter initiative requirements and will appear on the 2004 Alaska ballot. The initiative would eliminate civil and criminal penalties for people 21 years or older who "grow, use, sell or give away marijuana or other hemp products," while allowing subsequent laws to limit marijuana use in public.
http://www.adn.com/alaska_ap/story/4586606p-4557205c.html

Rehnquist report criticizes Feeney Amendment
In his annual year-end report on the federal judiciary, Chief Justice Rehnquist expressed his concerns about the Feeney Amendment, noting that it "could appear to be an unwarranted and ill-considered effort to intimidate individual judges in the performance of their judicial duties." The Feeney Amendment, a provision included in the AMBER Alert bill passed by Congress in Spring 2003, instructs the United States Sentencing Commission to take steps to reduce instances of "downward departure"-the term for sentences that fall below federally mandated minimums. The commission is directed to compile and maintain judge-by-judge records of sentencing departures and report to the attorney general, who in turn is obliged to provide the information to the Judiciary Committees of both houses.
http://www.cnn.com/2004/LAW/01/01/rehnquist.judiciary/

Drug Research News

Monitoring the Future
The University of Michigan's annual survey of 8th, 10th, and 12th grade students shows a decrease in illicit drug and cigarette use, with alcohol use remaining steady. The official press release highlights the second year of declining ecstasy (MDMA) use, with total use among students during the previous year dropping by half. The survey tracked illicit and licit drug use and attitudes among 48,500 students from 392 schools across the nation.
http://www.monitoringthefuture.org

Sativex nears approval
Britain's Home Office is expected this year to approve for prescription use Sativex, a new cannabis-based drug designed to alleviate symptoms of multiple sclerosis and severe neuropathic pain, the New York Times reported Jan. 27. Unlike Marinol, a currently available cannabinoid medication that uses sythentic delta-9-tetrahydrocannabinol (THC) as its active ingredient, Sativex contains all of the psychoactive agents in natural cannabis, but is designed with a 1:1 equal ratio of THC and cannabidiol (CBD). The new drug, developed by GW Pharamceuticals to be administered under the tongue, is expected to be less intoxicating than either cannabis smoking or Marinol.
http://www.cnn.com/2004/WORLD/europe/01/21/uk.cannabis/


Drug Policy Analysis Bulletin
c/o Federation of American Scientists
1717 K St., NW
Washington, DC 20036

FAS Homepage | Drug Policy | Index | Search | Join FAS