Current template: single.php

The Power of Salient Cost Estimates in the Federal Budget
Process: Credit Programs and Implications for Reform

Download this Post

Marvin Phaup1
Research Scholar and Professorial Lecturer
Trachtenberg School of Public Policy and Public Administration
The George Washington University
MPhaup@gwu.edu
August 5, 2019

DRAFT FOR COMMENT: Please do not quote without author’s permission.

Abstract: This paper describes a quasi-experiment with Congressional budgeting.
Results are consistent with the behavioral finding that salience of information in a
decision process often dominates its functional relevance. Specifically, I observe
two periods with radically different budgetary accounting for federal credit
assistance. In both periods, authoritative measures of costs are provided to
policymakers in highly salient form at key points of legislative decision. In the first
period, the estimate of opportunity cost is largely “noise,” with scant relevance to
the value of scarce fiscal resources consumed by the decision. By contrast, the
cost measure in the second period is a much closer approximation of total costs.
In both periods, minimization of cost, as measured, is strongly evident in
decisions. One implication for budget reform is that more functionally relevant,
salient measures of cost can improve the performance of public budgeting.
Timely introduction and maintenance of those measures, however, likely requires
an independent budgetary accounting authority.

INTRODUCTION

The current federal budget process–characterized by an unsustainable federal
debt trajectory, frequent lapses in budget adoption, inefficient and ineffective
program selection, a significant risk of annual government shutdowns, and a debt
ceiling more likely to produce default than fiscal stability–is widely regarded as
dysfunctional and in need of reform.2 Nonetheless, the existing process does
succeed in one important respect: it nudges policymakers to minimize the
reported budget cost of new legislation, conditional on addressing policy
objectives. This paper focuses on that success because it suggests a path to more
general, likely effective reforms.
The otherwise “broken” process promotes management of reported budget cost
by rendering cost estimates for new legislation highly salient at key legislative
points of decision.3 This process feature is a legacy of the Congressional Budget
Act of 1974 [CBA] ,4 [Sec .202] which directs the Congressional Budget Office
(CBO) to provide all information to the Congress that would assist in carrying out
the budgeting function. Accordingly, CBO releases a cost estimate for every bill
reported out of an authorizing committee. Congress–often indifferent to other
features of the CBA–has embraced fully the concept of cost minimization, given
its legislative policy objectives and current budgetary accounting.5 Members and
their staff often consult with CBO analysts to assure a minimum budget cost when
drafting legislation.

CREDIT ACCOUNTING: REGIME I

The power of salient cost information in affecting policy decisions is especially
evident in federal credit programs—direct and federally-guaranteed lending–
which have been subjected to significantly different budgetary accounting, or
“scoring” regimes, in the period since 1974. The first of those was a simple cash
flow measure: the effect of new loans on cash flows to and from the federal
government by budget year. Under that accounting, a $100 direct loan was shown
as increasing budget outlays and the deficit by $100 in the budget year the loan
was disbursed. In subsequent fiscal years, cash inflows were recorded when
repayments were received–or expected, for outyear budget projections. By
contrast, no federal cash flows were recognized at disbursement for federally
guaranteed loans advanced by third party, non-federal lenders. New guarantees
were “free” in reported budget cost until borrowers defaulted and the
government reimbursed lenders for losses. This accounting violated long-standing
budgetary accounting theory which holds that the total cost of a legislative or
other scorable action should be shown in the budget up-front, when the decision
is made to incur that cost (e.g. Office of Management and Budget. 2017)
For elected officials seeking to minimize the cost of credit assistance, the choice
between a direct loan and a guaranteed loan on the same terms to the same
borrowers was easy under cash flow accounting: the initial budget price was
either zero for a guarantee or the full amount of the loan for direct lending. In
fact, if the government collected an up-front guarantee fee, the budget year cost
was negative, indicating an inflow of fiscal resources, even if subsequent losses
were expected to exceed the fee. Similarly, Congress could score budget “savings”
5 An enduring Congressional proclivity for minimizing scored budget cost is documented in Meyers (1994) to pay-for other legislation by pairing a desired, but costly, policy proposal with
legislation to convert an existing direct loan program to a guarantee.
The effect on policy choice of the difference in cost estimates for equivalent
direct and guaranteed loans was dramatic (Figure 1). Specifically, the dollar
volume of new direct loans as a percent of the volume of new guarantees
declined from a peak of 80% in fiscal 1979 to 7% in 1992.
Congress made those policy choices even though it was aware that the cash-basis
cost measure was a poor proxy for fiscal cost and heavily biased toward
guarantees. For example, the 1967 Report (Ch.5) of the President Commission on
Budget Concepts had proposed changing the budgetary treatment of direct loans
to report long-term subsidy cost rather than annual cash flows. Further, the
Balanced Budget and Emergency Deficit Control Reaffirmation Act of 1987
directed the CBO to study and report recommendations for more accurately
measuring the cost of credit programs. Nonetheless, the salience of the cost
measure dominated its functional irrelevance in decisions.

CREDIT ACCOUNTING: REGIME II

CBO (1989) recommended that Congress adopt the expected, present value, longterm
loss on both loans and guarantees as the budget subsidy cost for credit
programs. It also recommended that this cost be appropriated before the loan
was obligated and recognized in budget outlays and the deficit when the loan was
disbursed. One objective of the reform was to assure that loans on the same
terms to the same borrowers would have comparable (equal) cost whether the
funds were advanced by the government or by a non-federal lender whose cost
recovery was guaranteed by the federal government (Phaup. 1985). Subsidy costs
were expected to be less than the amount advanced, unless the loan was a grant,
miscategorized as a loan (CBO 1991). Similarly, subsidy costs were expected to be
greater than zero, unless the loan was riskless and hence readily financed without
a federal guarantee by a private lender.
Congress enacted the Federal Credit Reform Act of 1990 (FCRA) as a part of the
Budget Enforcement Act of 1990. The budgetary cost of federal credit programs
specified in FCRA stopped short of the long-term loss recommended by CBO. It
differed from CBO’s recommendation by excluding two components of total cost.
The first, administrative costs, including loan servicing costs, were omitted to
avoid advance appropriation of wages, salaries and other operating expenses.
Second, the cost of market risk was excluded from subsidy cost, by mandating
that projected future cash flows from federal credit transactions be converted to
present values using interest rates on low risk Treasury securities of the same
maturity.6 Those omissions reduced average measured budget costs for credit by
5 – 10 percent of the amount loaned. (CBO. 2012, 2018).
FCRA corrected the bogus cost advantage of guarantees over direct loans, but
also changed fundamentally the legislative strategy for minimizing budget cost.
Instead of using loan guarantees rather than direct loans, policy makers and
analysts now had to consider the causal drivers of subsidy costs: fees and interest
rates charged borrowers; the frequency and severity of default; and the
government’s debt collection practices.
Figure 2 shows the distribution of initial estimates of subsidy rates for individual
direct loan programs for fiscal 1993. The federal policy portfolio consisted of 39
active direct loan programs, with a mean subsidy rate of about 23 percent, and
most rates clustered between 0 and 30 percent. Nine programs had subsidy rates
of 40 percent or higher.7

The new budgetary accounting for credit programs neither reduced the salience
of cost estimates nor diminished Congressional interest in minimizing reported
cost. Rather, the new practice of reporting subsidy rates as a percent of the
amount loaned by individual program heighten the visibility and importance of
cost estimates for federal credit policies.
Accordingly, with time and experience, Congress adapted to the new scoring by
increasing the use of direct loan programs and structuring legislation to reduce
net expected credit losses. By FY 2015, they had increased the number of direct
loan programs from 39 to 56 and reduced the mean program subsidy rate from 23
percent to 9 percent (Figure 3). Even more remarkably, Congress had succeeded
in “doing well, while doing good:” 25 direct loan programs reported negative
budget subsidy rates.8

The shift to negative rates transformed federal credit activity from costly to a net
source of fiscal resources that could be used to pay for other spending. For FY
2018, more than $600 billion in new direct loans and guarantees reported a net
budget gain of about $9 billion. 9

SPECIFIC POLICY ADJUSTMENTS UNDER REGIME II

Policymakers responded to the new scoring rule quickly. The George H. W. Bush
Administration proposed and secured a direct student loan demonstration project
in FY 1992, the first year FCRA became effective.10 President Clinton then
endorsed and signed into law—primarily on grounds of deficit reduction—the
Federal Direct Loan Program of 1993 whose estimated FRCA subsidy rate was
about half the rate for guaranteed student loans.11 To assure budget savings were
reflected in the cost estimate, the new legislation authorized the Secretary of
Education to require participating schools to switch from high cost guaranteed to
lower cost direct loans until 60 percent of new student loans were direct.
(McCann. 2010)12
In the Debt Collection Improvement Act of 1996 (DCIA), policymakers took steps
to reduce the subsidy cost of all federal-assisted credit by:

  • Barring the extension of new federal credit to borrowers who were more
    than 180 days delinquent on an existing federal direct or guaranteed loan;
  • Centralizing collection of defaulted loans in the US Treasury; and
  • Authorizing “offset” (collection) of non-tax debts owed the government
    including interest and penalties through reductions in amounts paid to
    defaulting borrowers for federal salaries, tax refunds, and Social Security
    benefits.

The Energy Policy Act of 2005 included a policy change aimed at reducing subsidy
costs for a specific credit program. Section 1703, Title XVII of the Act authorized
the Secretary of Energy to issue loan guarantees for projects that employ
innovative technologies at a fee set equal to the estimated subsidy cost of the
loan. This authority enables the Department to provide credit on below market
terms for target projects at a budget cost of zero.13
Reductions in credit subsidy costs were also used to pay for expansions of noncredit
spending. Examples include:

  • The Patient Protection and Affordable Care Act of 2010 was credited with
    5-year savings of $28 billion ($61 billion over 10-years) for a provision
    terminating the federal guaranteed student loan program and shifting all
    new federal student loans to the direct program, effective July 1, 2010.14
  • An increase in premiums and fees for federal mortgage guarantees offset
    $16.7 billion (5-year) and $35.7 billion (10-year) of the cost of extending a
    temporary payroll tax reduction, unemployment compensation, and
    deferral of a reduction in Medicare physician payments for one year,
    2012.15
  • The Budget Control Act of 2011, which established caps on discretionary
    spending through 2021 and created the failed Joint Select Committee on
    Deficit Reduction, also included an increase in federal Pell grant funding of
    $17 billion for FY 2012 and 2013. That increase was paid for ($9.6 billion
    over 5-years; $21.7 over 10-years) by two changes in the direct student
    loan program. The first changed a subsidized graduate student loan
    program into an unsubsidized, negative cost program by eliminating
    interest subsidies for in-school, post-school grace, and deferment periods.
    The second effectively raised origination fees for all students.16

 

SUCCESS! With Flies in the Ointment

On the face of it, FCRA looks to have been an unusually successful intervention in
a decision process. On closer examination, several flaws in the new accounting
regime sully the luster of achievement. First, as noted, the enacted measure of
subsidy cost omits some components of social cost so that budget costs
understate total costs on average by 5 – 10 percent of the amount loaned. This
understatement accounts for most of the negative subsidy rates reported by
credit programs under FCRA accounting. Moreover, the existence of “profitable”
credit assistance creates incentives for Congress and administering agencies to
increase issuance of loans, beyond the point that would maximize long-term wellbeing
for borrowers. One recent study found this development to be especially
problematic in the case of direct student loans and FHA mortgage guarantees
(Stanton, Rhinesmith, and Easterly. 2017). Direct student loans, with a negative
subsidy rate of 2.8 percent, accounted for over 85 percent of all direct loans
disbursed in FY 2018, while FHA mortgage guarantees–negative subsidy rate of 3
percent– made up nearly 50 percent of all new guarantees.17
Second, FCRA explicitly excluded federal insurance programs from coverage by
the new accounting. Thus, the same distortion of measured budget cost that
drew Members to loan guarantees, remains active for long term insurance in
which premiums are collected when coverage is extended but claims that trigger
outlays are made in future fiscal years. (CBO. 2005) In switching from Regime I to
Regime II, Congress gave up the opportunity to provide “free” loan guarantees
but gained the scoring option of “profitable” lending, while retaining the
favorable budgetary treatment of federal insurance.
Third, federal guarantees and insurance are not the only, nor the largest,
programs that provide benefits now but defer the recognition of cash-flow budget
cost to the distant, optimistically anticipated, heavily discounted future when
effective cost management is politically unfeasible. That distinction goes to social
insurance programs, including Social Security and Medicare.18 The budget costs of
those programs, as well as Federal pensions, other post-employment benefits,
and implied guarantees to government sponsored enterprises, are also almost
impossible to manage because they are recognized only when beneficiaries are
due to receive anticipated payments that they have been encouraged by
policymakers to expect (Phaup. 2009, 2019).

INTERESTING STORY: But Where is the Counter-Factual?

The case for a causal link between accounting and budget and policy choice would
be stronger with a well-specified model of legislative decisions affecting credit
policy that could provide a counter-factual path of “no FCRA treatment.” Absent a
general model–the V.O. Key (1940) lament—the current result may be regarded
as merely suggestive because of possible effects of variables, ignored in this
analysis, on credit policy. In fact, a robust model of budgetary decision-making
still eludes our professional grasp. That analytical gap has significantly retarded
efforts to improve the federal budget process. However, in this case, an inability
to isolate the policy effects of budgetary accounting from other factors is, at
most, a minor obstacle to reform. That is because the proposed accounting
changes are deduced from the objective of meeting user needs for cost
information so that trade-offs can be made with more functionally relevant
information. In that respect, the FCRA reform model is also informative.
Enactment of FCRA was motivated by the goal of improving budget decisions
through increasing the relevance of measured budget cost for credit. Budgeting is
a process of attempted constrained optimization in the face of scarcity. FCRA
simply redefined cost to measure more accurately the value of resources
consumed by specific policies providing credit assistance. The case for change
was developed from the purpose of the supported activity. Subsidy cost is
preferred to short-term cash flow because of its superior ability to inform budget
decisions. The possibility that the change might not improve decisions was and
remains insufficient to reject at least a trial of extending the new accounting
(Manzi. 2012).

IMPLICATIONS for REFORM

Congress is the architect of its budget process and its own budgetary accounting
standard setter. It is also subject to many of the same cognitive limitations as
other individuals and groups (Viscusi and Gayer.2015). Under current practice, we
observe persistent, functionally mis-measured costs, and inefficient program
choice; in behavioral terms: present and status quo biases.
12
An Independent budgetary accounting standards authority could improve policy
by defining cost measures more appropriate to the budget function without
restricting choice or compromising legislative sovereignty, while easing the
cognitive difficulty of the task (Hearn and Phaup. 2016). Consistent with past
practice, such an entity might be described as a permanent “Commission on
Budget Concepts.” Drawing on experience with the Federal Accounting Standards
Advisory Board (FASAB), the Commission would likely be more effective, if it were
a wholly non-federal entity (Patton and Mosso. 2009). That status would require
private funding and the absence of authority to impose its standards on Congress
or any other instrumentality of the federal government. In that case, standards
would be effectuated solely through the Commission’s ability to exert moral
suasion, its technical expertise, its non-partisan nature, and its public standing.19
As unlikely as it might seem, a behaviorally-informed case exists for better living
through better governmental budgetary accounting.

REFERENCES:

Arrow, Kenneth and Robert C. Lind (1970). “Uncertainty and the Evaluation of
Public Investment Decisions,” American Economic Review, 60, 3, 364-78
Bickley, James M. (2012). Subsidy Cost of Federal Credit: Cost to the Government
or Fair Value Cost? Congressional Research Service, Report R42503, April 25
https://fas.org/sgp/crs/misc/R42503.pdf
Burman, Leonard and Marvin Phaup (2012). “Tax Expenditures, the Size and
Efficiency of Government, and Implications for Budget Reform,” Tax Policy and
the Economy, 26, Jeffrey R. Brown (ed.), National Bureau of Economic Research,
University of Chicago Press, 92-124
Chetty, Raj, Adam Looney, and Kory Kroft (2009), “Salience and Taxation: Theory
and Evidence,” American Economic Review, 99, 4, 1145-77
19 If the Commission were certified by the American Institute of Certified Public Accountants as the source of
generally accepted budgetary accounting principles, then an audit body, such as GAO, might render an opinions as
to whether Presidential and Congressional budgets were consistent with those principles.
13
Chirico, Chad and Susanne Mehlman (2013), “FHA’s Single-Family Mortgage
Guarantee Program: Budgetary Cost or Savings?”, Congressional Budget Office,
October 21 https://www.cbo.gov/publication/44628
Congdon, William J., Jeffrey R. Kling, and Sendhil Mullainathan (2011), Policy and
Choice: Public Finance through the Lens of Behavioral Economics, Brookings
Institution, Washington, D.C.
Delisle, Jason and Jason Richwine (2014), “The Case for Fair Value Accounting,”
National Affairs, Fall https://www.nationalaffairs.com/publications/detail/thecase-
for-fair-value-accounting
Delisle, Jason (2017). Private in Name Only: Lessons from the Defunct Guaranteed
Student Loan Program, America Enterprise Institute, Washington, DC, February
Financial Economists Roundtable (2012). “Accounting for the Cost of Government
Credit Assistance,” Statement, October 16 https://fic.wharton.upenn.edu/wpcontent/
uploads/2017/01/FER-Statement-2012-10-16-12-final.pdf
Hearn, James J. and Marvin Phaup (2016). “Making Better Budget Decisions
Easier: Some Changes Suggested by Behavioral Research,” Papers on Reimagining
the Federal Budget Process, 5, Brookings Institution, Washington, DC, June 17
https://www.brookings.edu/wp-content/uploads/2016/07/Phaup-Hearn-Layout-
FINAL-PDF-1.pdf
Kahneman, Daniel (2011). Thinking, Fast and Slow, Farrar, Straus and Giroux, New
York.
Kamin, David (2013). “Risky Returns: Accounting for Risk in the Federal Budget,”
Indiana Law Journal, 88, 723-772 http://ilj.law.indiana.edu/articles/14-Kamin.pdf
Key, V.O, Jr. (1940). “The Lack of a Budgetary Theory,” American Political Science
Review 34,6, 1137 -1144.
Lucas, Deborah (2014). “Rebutting Arrow and Lind: why governments should use
market rates for discounting,” Journal of Natural Resources Policy Research, 6, 1,
85-91 http://gcfp.mit.edu/wpcontent/
uploads/2013/08/ArrowLindFinalPublished19390459.2013.874106-1.pdf
14
Lucas, Deborah and Damien Moore (2010). “Guaranteed versus Direct Lending:
The Case of Student Loans,” Measuring and Managing Federal Financial Risk, D.
Lucas, ed., National Bureau of Economic Research, University of Chicago Press,
163-211.
Lucas, Deborah and Marvin Phaup (2008). “Reforming Credit Reform,” Public
Budgeting & Finance, 28, 4 (Winter) 90-100.
Lucas, Deborah and Marvin Phaup (2010). “The Cost of Risk to the Government
and Its Implications for Federal Budgeting,” Measuring and Managing Federal
Financial Risk, D. Lucas, ed., National Bureau of Economic Research, University of
Chicago Press, pp 29 -54.
Mankiw, N. Gregory (2016). Principles of Microeconomics, Eighth Edition, Cengage
Learning, Boston, Chapter 10.
Manzi, Jim (2012). Uncontrolled: The Surprising Payoff of Trial-and-Error for
Business, Politics, and Society, Basic Books, New York.
McCann, Clare (2010). “Student Loan Program – History,” New America EdCentral
http://www.edcentral.org/edcyclopedia/federal-student-loan-programs-history/
Meyers, Roy T. (1994). Strategic Budgeting, University of Michigan Press. Ann
Arbor.
Office of Management and Budget (2017). Budget of the U.S. Government,
Analytical Perspectives, FY 2018. Chaapter 8, “Budget Concepts.”
https://www.whitehouse.gov/sites/whitehouse.gov/files/omb/budget/fy2018/sp
ec.pdf
Patton, James and David Mosso (2009). “Is the Federal Accounting Standards
Advisory Board Independent? Can FASAB be More Independent?” Journal of
Government Financial Management, Fall, 58,3, pp.61-63.
Phaup, Marvin (1985). “Accounting for Federal Credit: A Better Way,” Public
Budgeting & Finance, 5, 3 (September), pp. 29-39
Phaup, Marvin (1996). “Credit Reform, Negative Subsidies, and FHA,” Public
Budgeting & Finance, 16, 1 (March), pp. 23-36
15
Phaup, Marvin (2009). “Federal Use of Implied Guarantees: Some Preliminary
Lessons from the Current Financial Distress, Public Administration Review, 69, 4
(July/August) 651-659. http://onlinelibrary.wiley.com/doi/10.1111/j.1540-
6210.2009.02014.x/abstract
Phaup, Marvin (2019). “Budgeting for Mandatory Spending: Prologue to Reform,”
Public Budgeting & Finance, 39 ,1 (Spring) pp. 24-44.
Stanton, Thomas, Alan Rhinesmith and Michael Easterly (2017). “Federal Credit
Programs: Borrower Outcomes Matter More than Volume,” http://thomasstanton.
com/wp-content/uploads/2017/06/Federal-Credit-Programs-borroweroutcomes-
as-a-focus.pdf
Thaler, Richard and Cass Sunstein (2008). Nudge: Improving Decisions about
Health, Wealth, and Happiness. New Haven, CT: Yale University Press.
U.S. Congressional Budget Office (1989). Credit Reform: Comparable Budget Costs
for Cash and Credit, CBO Study, December
U.S. Congressional Budget Office (1991). “An Explanation of the Budget Changes
Under the Credit Reform,” Staff Memorandum, April
https://www.cbo.gov/sites/default/files/102nd-congress-1991-
1992/reports/7916.pdf
U.S. Congressional Budget Office (2004). Estimating the Value of Subsidies for
Federal Loans and Loan Guarantees, CBO Study, August
https://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/57xx/doc5751/08-19-
creditsubsidies.pdf
U.S. Congressional Budget Office (2005). The Risk Exposure of the Pension
Benefit Guaranty Corporation, September
https://www.cbo.gov/sites/default/files/cbofiles/ftpdocs/66xx/doc6646/09-
15-pbgc.pdf
U.S. Congressional Budget Office (2009). Letter to Senator Gregg, July 27, 2009
https://www.cbo.gov/sites/default/files/111th-congress-2009-
2010/reports/gregg_studentloans__09-07-27.pdf
U.S. Congressional Budget Office (2012). Fair Value Estimates of the Cost of
Federal Credit Programs in 2013, June
16
U.S. Congressional Budget Office (2019). Fair Value Estimates of the Cost of
Federal Credit Programs in 2020, May https://www.cbo.gov/system/files/115thcongress-
2017-2018/reports/54095-2019fairvalueestimates.pdf
U.S. Government Accountability Office (2016). “Credit Reform: Current Method to
Estimate Credit Subsidy Costs is More Appropriate for Budget Estimates than a
Fair Value Approach,” GAO-16-41. January. www.gao.gov/assets/680/674905.pdf
Viscusi, Kip and Ted Gayer. 2015. “Behavioral Public Choice: The Behavioral
Paradox of Government Policy,” Harvard Journal of Law and Public Policy, 38: 973
– 1007.
SEA 2019 – Phaup –  Power of Salient Costs Estimates

One thought on “SEA 2019 – Phaup –  Power of Salient Costs Estimates

  • Avatar
    June 23, 2020 at 7:09 AM
    Permalink

    You Want $5000 Or More Each Month Working With Local Businesses ?

    Google Local Raider is a complete Google Guarantee business-in-a-box.
    Everything you need to start earning passive monthly commission check is provided in a done-for-you format including a detailed,
    no-fluff guide that walks through the entire process step by step.

    MORE INFO HERE=> https://zeep.ly/Lqic4

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.