KEEPING SCORE: THE FRAILTIES OF THE FEDERAL DRUG BUDGET

Total federal government expenditures for antidrug activities have become a
centerpiece in the national debate on drug policy. The rapid growth in that
total was a major indicator of the extent of federal government commitment
to dealing with what was perceived to be the most prominent social problem
in the late 1980s. Even more importantly, the allocation of that
budget--between supply-control programs on the one hand and treatment and
prevention on the other--is perhaps the most widely cited piece of evidence
that the United States is committed primarily to law enforcement as the
principal element of its drug policy. Given the prominent role that federal
budget figures have come to play in the policy debate, it is noteworthy
that few have paid any attention to their origins.

This issue paper explores the history of the federal drug-control budget as
well as the methodology used to produce the numbers in it. It argues that
what is commonly referred to as the "federal drug control budget" is not a
budget in the textbook sense: that is, it is limited in its capacity to
either ensure accountability or establish policy priorities. At best, the
federal drug budget provides a descriptive level of effort. And even as a
historical description, the numbers may misrepresent the total level of
expenditures as well as their distribution across programs to control both
supply and demand.

This paper begins with a brief review of the federal drug-control
expenditures and the history of the drug budget. It then examines the
differing methods used to calculate agency drug-control budget estimates
and the consequences of the differences. Such "scorekeeping" varies from
agency to agency and produces a distinctly ad hoc collection of estimated
drug-control expenditures. Finally, the paper argues that the federal drug
budget better serves as an advocacy tool than as a mechanism to drive
policy changes.

History of the Drug-Control Budget

Drug control includes a broad spectrum of programs at the federal level.
Supply-reduction (or law enforcement) programs range from international
efforts in drug-producing countries, to efforts to interdict smuggled drugs
as they cross U.S. borders, to the support of local law enforcement
efforts. Demand-reduction expenditures primarily consist of drug treatment
and prevention grants to state and local jurisdictions as well as the
provision of treatment services to veterans. At least seven cabinet
agencies have significant drug-control programs (Defense, Education, Health
and Human Services, Justice, State, Transportation, and Treasury).

The practice of calculating estimates of total federal drug-control
expenditures dates back to the late 1970s.[1] According to a former Office
of Management and Budget (OMB) official, a member of the then Office of
Drug Abuse Policy (within the Executive Office of the President) would
canvass the agencies and request estimates of their antidrug expenditures.
The White House then would cite these figures in supporting documents
released with proposed policy initiatives. This collection of estimates,
however, served little if any control function; i.e., it was not used to
monitor agency performance or to set the next-year budget figures.[2]

By the middle of the 1980s the abuse of illicit drugs had risen high on the
domestic policy agenda. In response to increasing public attention to the
drug issue and a growing number of requests for estimates on drug-control
expenditures, OMB took on the task of tabulating the drug budget. OMB
sought to introduce a more systematic method of collecting and updating
figures on federal programs. In 1985, OMB asked each of the agencies it
thought to have a role in the so-called war on drugs to estimate its annual
drug-control expenditures dating back to FY 1981 (the last budget year of
the Carter administration). The changes in 1985 marked the beginning of a
continuing effort to refine the estimates that represented antidrug
programs. The Office of Management and Budget (eventually joined by the
Office of National Drug Control Policy (ONDCP)) would make alterations to
the tables in an attempt to produce a consistent set of numbers--over time
and across categories.

The Kalder Tables (named after the examiner who maintained them) served a
very useful function for a small group of Washington policymakers and their
staff. The tables provided a single source for figures on the drug-control
budget that would be extremely difficult for an individual to compile.
Figures from the tables were used to give answers to various inquiries, and
portions also appeared in press releases and periodic reports. OMB also
informally distributed updated versions of the tables to a collection of
approximately 50 interested executive agencies, policy-coordinating groups,
and congressional staff members. While OMB was producing the document,
however, the tables continued to serve as a reference document, and they
had only a limited effect on programming and budget decisions; certainly
the allocation across program areas did not play a significant role in debate.

The creation of the position of Drug Czar--the director of ONDCP--marked a
new phase in the federal drug budget's history. The 1988 act that created
the ONDCP also required the new office to compile a federal drug-control
budget request and to certify agency requests as part of developing a
national strategy (PL 100-690, section 1003). OMB's Kalder Tables became
the basis for ONDCP's budget tracking and reporting system, but the
political stakes surrounding drug-control expenditures had been raised.
Members of Congress expected ONDCP's budgetary responsibility to play a
critical role in coordinating federal programs.

With this rise in political intensity, budget analysts in the executive
branch began to pay more attention to which agencies were included in the
tables and how the estimates were calculated. The Health Care Financing
Administration (HCFA), which administers Medicare and Medicaid, was added,
for example. [3] Other agencies, such as the Secret Service, saw an
association with the war on drugs to be politically expedient and sought
inclusion. Still other departments and bureaus revised their methodology
for calculating their estimates in an attempt to provide what they
considered a more accurate reflection of their level of effort.

The type of tabulation that the drug budget represents has a long history
in federal budgeting. The publication of the federal drug budget by the
ONDCP represents only a recent example of special analyses that the
Executive Office of the President has produced for over four decades. In
the jargon of the budget analyst, these tables often are "cross-cuts"--a
collection of expenditure estimates that span several departments and
agencies. The Office of Economic Opportunity collected similar estimates in
conjunction with the war on poverty. The OMB currently assembles budget
estimates for other cross-cutting efforts such as AIDS programs and
research and development expenditures. Unlike other cross-cuts, however,
the numbers that comprise the federal drug budget have become much more
salient relative to policy discussions.

The role played by the drug budget in the policy debate has revolved around
two questions. First, the public focused on the question, "Is the federal
government doing `enough' in response to the threat posed by illicit
drugs?" Elected officials highlighted the growth in resources devoted to
antidrug programs as evidence of their action. Citing the growing budget
was particularly popular in the mid-1980s, as estimates of the total number
of drug users continued to climb. Based on official estimates, the
budgetary increases were significant. Antidrug expenditures quadrupled in
real terms from FY 1981 to FY 1992 (Table 1), with annual growth averaging
about 15 percent.[4]

(Picture)

In the early 1990s, however, a second question emerged, enabling the drug
budget to maintain a prominent role in the policy discussion. The debate
shifted from "Is the federal government doing enough?" to "Is the federal
government doing the right things?" In short, the issue was whether the
program emphasis should be placed on drug law enforcement (i.e.,
supply-side programs) or on prevention and treatment efforts (i.e., the
demand side).

The distribution of resources as measured in the federal drug budget--the
supply/demand split--became the metric for the debate. In contrast to the
dramatic growth in the total level of resources, the composition of the
federal drug-control budget exhibited a more gradual change. The proportion
of resources devoted to supply-side efforts rose from 59 percent in FY 1981
to a peak of 71 percent in FY 1986. Since then, the share of resources
devoted to law enforcement activities has remained fairly constant,
representing just over two-thirds of the total.[5]

Advocates of increased funding for treatment and prevention programs argue
that this distribution illustrates the federal government's overemphasis on
the supply-side activities. They contend that a 50/50 split would be more
appropriate. Defenders of the current composition of funds point to the
federal government's unique responsibility to protect the borders and wage
the war on drugs overseas. Both camps often use the federal drug budget as
the scorecard to count up victories or defeats.

What began as an ad hoc collection of numbers for inclusion in press
releases took on a life of its own. The drug budget shifted from a
peripheral supporting role to the center of policy deliberations. Implied
was the notion that these tables should be used as the mechanism to drive
policy changes. Yet the drug budget in many cases remained a compilation of
"best guesses." An examination of the methodology used to arrive at those
numbers raises serious questions as to whether the drug budget can bear the
weight that the public debate has placed upon it.

Scorekeeping

The methods used to calculate the "drug portion" of an agency's budget vary
considerably, but they generally fall into one of three types. Type 1 is
the most straightforward; in it, all agency funds are considered to be
drug-related. These agencies present total budget accounts whose levels are
specifically stated in appropriation bills and can be tracked by the
various executive and legislative accounting systems. For example, the
budgets for the Drug Enforcement Administration (DEA), the Organized Crime
Drug Enforcement Task Forces (OCDETF), and the Bureau of International
Narcotics Matters (INM) in the Department of State are "scored" as 100
percent drug-related (ONDCP, 1992, pp. 90, 116, 146). DEA, OCDETF, and INM
prove to be the exception, however. These Type 1 agencies account for only
16 percent of the FY 1992 total.

For the majority of agencies, however, it is not possible to use account
totals to estimate the level of drug expenditures. For example, in law
enforcement agencies with multiple missions, drug-control efforts may
represent only a portion of their total activities. In these cases, Type 2
or a "flat percentage" scorekeeping is used. The agency calculates its
drug-control budget as the product of its total operating budget and an
estimated percentage of its time spent on drug-control activities. ONDCP
explains that the Coast Guard's drug-control budget represents "estimates
of time spent by Coast Guard operational facilities (boats, cutters, and
aircraft) in the performance of drug-related missions" (ONDCP, 1992, p.
157). This estimate is reported to be 19 percent of the Coast Guard's total
operating time. Type 2 agencies account for about one-quarter of the total
drug budget.

Although the Type 2 methodology is relatively transparent, it is not
without problems. For example, the 19 percent figure from the Coast Guard
example above was determined in the late 1980s and has remained
approximately the same since. But the Coast Guard appears to have changed
its priorities, at least according to its rhetoric. The emphasis has
shifted from the agency's role in stopping drug smugglers and toward a
concern for environmental protection and illegal immigration. The 19
percent figure remained until the FY 1994 budget, however, as there was
little incentive to change it.

The final scorekeeping category, Type 3, is a catchall for all the
remaining methodologies, often combining features of Types 1 and 2. Type 3
agencies account for almost 60 percent of the total federal drug budget.
For a few agencies, the drug budget calculation produces methodologies that
some may describe as algorithms, others as alchemy. The Department of
Veterans Affairs (DVA) is one such example. ONDCP's explanation of how DVA
estimates its drug budget is as follows:

"The drug percentage represents the drug treatment costs for all primary
and secondary drug diagnoses in all hospital bed sections, including costs
of specialized drug dependence treatment units which account for
approximately one-third of total treatment costs. The drug portion of
medical care costs is broken down into four general components: 100 percent
of the medical costs of patients participating in drug treatment programs;
100 percent of the medical costs of patients with a primary diagnosis of
drug abuse but who are not participating in drug treatment programs; 50
percent of the costs of patients with a secondary diagnosis of drug abuse;
25 percent of the costs of patients with a secondary diagnosis of substance
abuse. Costs for drug treatment programs are counted at 100 percent. The
percentage of costs attributable to the treatment of patients with drug use
disorders in other specialized treatment programs was calculated to be 33.5
percent" (ONDCP, 1992, p. 194).

The DVA example reveals that how much of an agency's budget is attributed
to the antidrug effort is the product of a number of subjective decisions.
This is not an attempt to argue that the methods used here are right or
wrong, only to point out that a number of judgments are made in the
scorekeeping of funds in Type 3 agencies.

The opaque nature of the Type 3 methodologies proves somewhat problematic
to those seeking to use the drug budget to direct policy changes. The
problem is further complicated when the methodologies change over time. The
Department of Defense, supposedly spending over $1 billion for drug
control, has shown considerable opportunism in its allocation; for example,
in FY 1991 it attempted to move the costs of its Over-the-Horizon
Backscatter radar, originally designed to detect incoming Soviet bombers,
to its drug-control budget, a bit of budgetary legerdemain involving $214
million. Congress objected to this, but there remains considerable question
as to whether the DoD drug expenditure figure is not highly inflated.

The Department of Defense was not the only agency to seize a political
opportunity in the process of estimating its contribution to the antidrug
effort. The Secret Service also attempted to inflate its estimate when it
revised its methodology in 1991. In its submission to ONDCP, the agency
claimed that approximately one-quarter of its total $400 million budget
represented drug-control expenditures, whereas prior-year claims had been
less than $5 million. Some of its claimed contributions were clearly
tangential at best. For example, the agency included in its original
estimate the cost of protecting former First Lady Betty Ford when she
traveled to give drug prevention lectures. Analysts at both OMB and ONDCP
objected to the inclusion of such costs, and the three finally agreed that
about 10 percent of the Secret Service's total budget was drug-related.

Evidence of other methodological shifts appears in Table 2. This table
gives, for three successive years of publication, the estimated
drug-control expenditures of selected departments for the same year, FY
1991. Each column represents a different annual publication by the ONDCP.[6]

(Picture)

One example of how changes in the scorekeeping methodology can produce
significant changes in drug budget estimates emerges from the FY 1993
budget request for the Alcohol, Drug Abuse, and Mental Health
Administration (ADAMHA). ADAMHA administers the block grant that provides
funds to the states for the purpose of treating and preventing substance
abuse--which includes both alcohol and illicit drugs. In the FY 1992
budget, the "drug portion" of these grants represented approximately 40
percent of the total (ONDCP, 1991, p. 154). In the FY 1993 budget, however,
that percentage was raised to 55 to include a portion of the resources
devoted to clients with a primary diagnosis of alcohol abuse but a
secondary diagnosis of abuse of illicit drugs and for the treatment of
alcohol abusers under the age of 21 since they are prohibited from
consuming alcohol up to that age (ONDCP, 1992, p. 46). The result is a
demand reduction total that is approximately $200 million higher than it
might otherwise have been.

The timing of this change, occurring when the Bush administration was under
increasing pressure to increase treatment and prevention funding, can give
rise to some skepticism. In fact, the methodological changes for the
departments listed in Table 2 had a significant effect on the distribution
of resources between supply and demand efforts. These scorekeeping changes
resulted in an increase for the two treatment and prevention agencies (HHS
and Veterans) of $458 million (+24 percent), while the two predominantly
law enforcement departments (Justice and Defense) had a net decrease of $41
million (-1 percent).

It is important to note that in both the Defense and HHS cases the
administration disclosed the changes in the supporting documents that
accompanied the release of the budget proposals.[7] In neither case was
there any obvious effort to hide the methodological shifts, nor were they
highlighted. Although one may find fault in the rationale used to justify a
change, such arguments can rarely proceed beyond the stage of reasonable
individuals agreeing to differ.

As noted earlier, the potential to manipulate these estimates is
problematic for those who frame the drug policy debate in terms of the
distribution of resources between the supply and demand reduction programs.
The estimates prove quite sensitive to changes in the underlying
assumptions. For example, it is possible to raise the percentage of
expenditures devoted to law enforcement activities, currently at 68
percent, to 75 percent under a new set of assumptions. In a second
scenario, this figure drops to 55 percent.[8] Neither case would involve
changes in the actual appropriation of funds.

In times of constrained resources, the incentive to manipulate can become
particularly strong. The murkiness of these calculations enables elected
officials to claim credit for increased antidrug funding without adding to
the deficit. The political benefit is twofold. The demand reduction
percentage goes up, but such an "increase" does not involve the spending of
new money, thus having no effect on the deficit. Nor is any money
subtracted from another program account with political salience.

Other Limitations

The potential for manipulation of the estimates making up the federal drug
budget is not the only condition that limits its utility as a mechanism to
drive policy changes. One constraint is the decentralized nature of the
budget process. During the early stages of putting together the President's
budget request, the ONDCP collects individual submissions from over 50
agencies. At this point in the process it is important to note what the
ONDCP cannot do. The office does not have discretion over a set amount of
money to distribute to the various agencies. Instead, the ONDCP's review of
each agency's drug-control efforts takes place in the context of the larger
budget process. Agencies must weigh increased funding for antidrug efforts
that may come at the expense of additional resources for other programs.[9]
As a result, ONDCP can become an advocate for funding increases that the
potential recipient opposes.

During the last three years of the Bush administration, for example, ONDCP
sought increased drug treatment funding over the objections of HHS. That
department opposed the additional funds on the grounds that it had other
priorities it would prefer to fund first. In other words, HHS would have
preferred to have the marginal dollar spent on child immunization or Head
Start before additional funds were devoted to drug treatment. The budget
process within the executive branch, then, is more a negotiation than a
central authority issuing directives.

The fragmented nature of the budget process continues on Capitol Hill,
where the drug budget falls under the jurisdiction of nine different
appropriations bills. Most funding decisions are made at the subcommittee
level. Funding for the Drug Enforcement Administration, for example, falls
under the jurisdiction of one committee, whereas a different group of
appropriators makes decisions on the Education Department's drug prevention
grants. Given the lack of centralization, the image of reordering the
federal government's drug-control priorities by taking dollars from law
enforcement and giving them to demand reduction programs becomes something
of a polite fiction.

A final limitation of using the federal drug budget to direct U.S. drug
policy is that the federal government represents only a minority share of
the resources devoted to drug-control efforts. If what is being sought is a
representation of the nation's drug-control efforts, one has to look beyond
the federal level. Table 3 provides estimates for FY 1990 of drug-control
funding for all levels of government.

(Picture)

Based on some admittedly rough calculations, the federal government
probably accounts for about one-third of over $30 billion of public funds
devoted to antidrug programs. And the distribution between supply and
demand reduction programs is even more skewed at the state and local level.
For each dollar they spend on treatment and prevention, state and local
governments devote about five dollars to law enforcement.

Conclusions

What can be concluded about the federal drug budget is that there are
limits to the functions it can serve. As an advocacy tool, it serves the
interests of both proponents and opponents of the current level and mix of
federal programs. But as a mechanism to change policy, the drug budget
proves to be a fragile construction. The algorithms used to calculate many
agency expenditures are opaque constructions and vulnerable to
manipulation. Consequently, the official figures may overstate the federal
government's expenditures on antidrug activities, and the distribution
between supply and demand resources is better thought of as a fairly broad
range rather than a precise point. Using the federal drug budget to direct
national policy is further limited by structural constraints. The
fragmented nature of the budget process and the significant role played by
state and local governments limit the drug budget's utility for policymaking.

The debate over policy priorities may continue to focus on the distribution
of resources between federal supply and demand reduction programs. If the
discussion continues to be framed in these terms, however, little progress
can be expected. Moreover, the drug problem as an issue appears to be
moving lower on the national domestic policy agenda. Such a movement is
unfortunate from the standpoint that the problems associated with illicit
drug use have far from disappeared. But a lessening of national political
interest in drug issues could prove advantageous if questions about the
federal drug budget move during this period toward the periphery of the
larger policy debate. Perhaps the debate over the relative distribution of
an artificial set of numbers can be displaced by a new focus on identifying
specific programs that address the harms associated with illicit drug use.

References

Office of National Drug Control Policy, National Drug Control Strategy,
Budget Summary, Washington, D.C. (annual).

Strategy Council on Drug Abuse, Executive Office of the President, Federal
Strategy for Drug Abuse and Drug Traffic Prevention, Washington, D.C., U.S.
Government Printing Office, 1979.

[1] See Strategy Council on Drug Abuse, Executive Office of the President
(1979).

[2] The Office of Drug Abuse Policy did send the figures to the OMB for
Rverification.S Individual OMB examiners typically would sign off on the
figures as a matter of course, having had little incentive to make further
inquiry.

[3] Estimates of drug-control expenditures from the 1970s did include
figures for Medicaid and Medicare. Such figures did not appear in the
tables prepared during the Reagan administration, however.

[4] The pattern of growth has been far from smooth, however. Annual changes
in real drug expenditures have fluctuated from a high of 64 percent (FY
1986/87) to a low of P5 percent (FY 1987/88).

[5] The stability in the distribution of resources is particularly
interesting given that in 1986 the Congress passed a resolution that any
additional resources for the war on drugs be divided equally between supply
and demand programs.

[6] Table 2 also suggests that the DVA underwent a similar methodology
shift. It is unclear what the rationale was for this change.

[7] In the FY 1994 set of drug budget estimates published by the OMB,
however, the methodologies were not disclosed. As a result, it is not
possible to determine what changes were made to the Department of Veterans
Affairs estimate.

[8] The changed assumptions necessary to produce these kind of shifts in
the supply/demand split prove to be quite modest. For example, ONDCP (1992,
p. 53) reports that the Health Care Financing Administration will spend
approximately $200 million in FY 1992. This figure represents 0.1 percent
of HCFAUs total expenditures of over $190 billion. If a rationale could be
developed to raise the estimate to 0.2 percent, the administration could
claim an additional $200 million for treatment programsQan 11 percent
increase over the FY 1992 level for the federal treatment budget total.

[9] Budget deficit politics in general produce this type of zero-sum
calculation. More recent attempts to cap domestic discretionary spending
through legislation have increased the need to examine budgetary tradeoffs
within and across departments.

This research was sponsored by RAND's Drug Policy Research Center and
funded by The Ford Foundation.

RAND issue papers explore topics of interest to the policymaking community,
with the intent of stimulating discussion in a policy area. Although issue
papers are formally reviewed, authors have substantial latitude to express
provocative views without doing full justice to other perspectives. The
views and conclusions expressed in issue papers are those of the authors
and do not necessarily represent those of RAND's research sponsors. RAND is
a nonprofit institution that seeks to improve public policy through
research and analysis. Results of specific studies are documented in other
RAND publications and in professional journal articles and books. To obtain
information about RAND studies or to order documents, call Distribution
Services, 310 451-7002, or access RAND's URL, http://www.rand.org/

Checked-by: Richard Lake