Congressional Budget Office
Director's Blog

Yesterday CBO and the staff of the Joint Committee on Taxation (JCT) completed a preliminary estimate of the direct spending and revenue effects of the reconciliation proposal that represents one component of the health care legislation being considered by the Congress. The other component is a bill, H.R. 3590, that the Senate passed in December.

CBO and JCT estimate that enacting the combination of the reconciliation proposal and H.R. 3590 as passed by the Senate would reduce federal budget deficits by $138 billion between 2010 and 2019. Although CBO does not generally provide cost estimates beyond the 10-year budget projection period, many Members have requested our analysis of the long-term budgetary impact of broad changes in the nation’s health care and health insurance systems. Therefore, we have developed a rough outlook for the decade following the 2010-2019 period. We estimate that the combined effect of enacting those two pieces of legislation would be to reduce federal budget deficits during that following decade relative to those projected under current law—with a total effect that is in a broad range around one-half percent of gross domestic product (GDP). That calculation is very uncertain, and the imprecision of the estimate is intended to reflect that uncertainty.

Many people have raised questions about those projections of deficit reduction and similar projections that we have made regarding earlier pieces of health legislation. Their questions often focus either on the uncertainties surrounding the various technical, behavioral, and economic factors underlying the estimates, or on uncertainties as to whether the legislation would ultimately be implemented as written.

Some analysts believe that CBO is underestimating the ultimate costs of the new subsidies to buy health insurance (which could make the legislation deficit-increasing instead of deficit-reducing). Others assert that CBO is underestimating the ultimate savings from changes in the Medicare program (which could make the legislation reduce deficits by more than we have estimated). Certainly, the budgetary impact of broad changes in the nation’s health care and health insurance systems is very uncertain. However, CBO staff, in consultation with outside experts, has devoted a great deal of care and effort to this analysis, and the agency strives to develop estimates that reflect the middle of the distribution of possible outcomes. As a result, we believe that CBO’s estimates of the net savings that would result from the legislation have a roughly equal chance of turning out to be too high or too low.

Focusing on another area of concern, some observers argue that CBO’s estimates are unrealistic because the Congress will not allow the Medicare spending cuts and future tax increases in the proposals to take effect. CBO’s responsibility to the Congress is to estimate the effects of proposals as written and not to forecast future legislation. However, the agency does try to provide information about the consequences of implementing proposals. For example, our cost estimate for the bill taken up by the Senate in December and our estimate for the House bill last October noted that inflation-adjusted Medicare spending per beneficiary would slow sharply under those proposals. We estimated that growth in such spending under the Senate bill would drop from about 4 percent per year for the past two decades to roughly 2 percent per year for the next two decades; whether such a reduction could be achieved through greater efficiencies in the delivery of health care or would reduce access to care or diminish the quality of care is unclear. In addition, CBO’s estimates have shown that relaxing previously enacted constraints on Medicare spending can add significantly to long-run budget deficits, as we wrote in answer to a question last fall about the effects of combining the House bill with a change in the so-called Sustainable Growth Rate mechanism for Medicare’s payments to physicians.

As we reported yesterday, budget deficits would be reduced, in our estimation, if the reconciliation proposal and Senate-passed health bill are enacted and remain unchanged throughout the next two decades.  However, the legislation would maintain and put into effect a number of provisions that might be difficult to sustain over a long period of time. Whether any of its provisions—and if so, which ones—might be changed in the future is not for CBO to judge.

Today, in a letter responding to questions from Congressman Ryan, CBO described the effects on the federal budget of enacting the reconciliation proposal and the Senate-passed health bill if:

• The excise tax on insurance plans with relatively high premiums—which would take effect in 2018 and for which the thresholds would be indexed at a lower rate beginning in 2020—was never implemented;
• The annual indexing provisions for premium subsidies offered through the insurance exchanges continued in the same way after 2018 as before—in contrast with the reconciliation proposal, which would slow the growth of subsidies after 2018;
• The adjustment to physician payment rates under Medicare that was passed by the House last fall was included; and
• The Independent Payment Advisory Board—which would be required, under certain circumstances, to recommend changes to the Medicare program to limit the rate of growth in that program’s spending, and whose recommendations would go into effect automatically unless blocked by subsequent legislative action—was never implemented.

We estimated that if this set of changes was made, the legislation as modified would increase federal budget deficits during the decade beyond 2019 relative to those projected under current law—with a total effect during that decade in a broad range around one-quarter percent of GDP.

One question receiving much attention in the discussion of the pending health care legislation is:  How would it affect unauthorized immigrants?  In a recent letter (dated October 29), we noted that unauthorized immigrants would constitute “about one-third” of the 18 million nonelderly residents who we estimate would remain uninsured under H.R. 3962, the bill currently being considered in the House. In our preliminary analysis of an earlier version of the legislation, H.R. 3200 as introduced, we said that “nearly half” of the 17 million residents who we estimated would remain uninsured would be unauthorized immigrants.

The use of the terms “about one-third” and “nearly half” was meant to convey the uncertainty and imprecision surrounding our estimates of the characteristics of the remaining uninsured population. Because of that uncertainty and imprecision, we cannot provide a specific figure for coverage of unauthorized immigrants under any of the proposals. Despite the difference in wording, we would not expect any significant differences between the two bills in the number of uninsured who are unauthorized immigrants, because the relevant features of the two proposals are similar. (Our analysis of H.R. 3200 was preliminary and based on specifications rather than a reading of the legislative language.) Further, we have not changed our methodology for estimating the relevant factors in the intervening period.

This morning I testified before the Senate Finance Committee on CBO’s two major health reports, Key Issues in Analyzing Major Health Insurance Proposals, and Budget Options for Health Care. The first examines the principal elements of reform plans that would affect our estimates of the effect of such plans on federal costs, insurance coverage, and other outcomes. The second comprises 115 discrete options to alter federal programs, affect the private health insurance market, or both. These two reports are the culmination of a tremendous effort by many people at CBO as we’ve redoubled our capacity to analyze these complex issues. (I was recently asked to testify about these volumes before the Senate Budget Committee; click here for the blog summarizing my testimony there.) 

As I emphasized today in my statement before the committee, health care reform is an urgent issue. In contrast with the situation in the economy and financial markets, our system for delivering and paying for health care is not fundamentally different this year from last year. However, the relatively gradual pace of change in health care is rarely seen as an argument for deferring action. Our current health system evolved over years and decades, and while coverage could be substantially expanded in a few years, it could take many years or even decades for the thoroughgoing changes needed to improve the system’s efficiency to come fully to fruition. Because of the lead times involved in realizing these efficiencies, nearly all analysts think that those changes should begin now.  

Yesterday’s testimony on the effects of recent financial market turmoil on pension assets has generated a significant amount of interest, so I wanted to follow up on one topic: overinvestment in an employer’s stock.

Many participants in retirement plans appear to be taking on unnecessary risk by investing in individual stocks rather than a diversified portfolio. The result is that those workers assume excessive risk for which they do not receive a higher expected return. (Those workers may feel they have inside information or insights that will allow them to outperform the market with particular investment choices, but the evidence suggests that unless you’re Warren Buffett, trying to outguess the market usually doesn’t work.)  Investing excessively in one stock that also happens to be your employer’s stock is even riskier — if the company runs into trouble, both your retirement assets and your job may be in danger.

Despite those risks (again, for which workers don’t receive higher expected returns on their investments, on average), a significant number of 401(k) participants hold the bulk of their assets in company stock. According to calculations by the Employee Benefits Research Institute, 47 percent of 401(k) participants were enrolled in plans that offered company stock as an option as of the end of 2006. Of those participants, 7.3 percent held more than 90 percent of their assets in company stock, and over 15 percent held over half their assets in company stock. (See page 33 of EBRI Issue Brief 308, “401(k) Asset Allocation, Account Balances, and Loan Activity in 2006.”) At yesterday’s hearing, I didn’t make clear that the 7.3 percent figure applied only to those who were in plans offering company stock. So the overall share of 401(k) participants with 90 percent or more of their assets invested in company stock is more like .47*7.3=3.4 percent. It’s still too high.

The good news is that the trend is towards less investment in company stock. For example, in 1999 EBRI estimated that 19.1 percent of all 401(k) assets were held in company stock. (See figure 20 on page 26 of EBRI issue brief 308.) By 2006, that share had fallen to 11.1 percent, undoubtedly driven in part by the example of the collapse of Enron and other corporate failures. In addition, provisions of the Pension Protection Act of 2006 limit the amount of time that an employer may insist participants keep assets in company stock.

Despite these auspicious trends, there remains substantial scope to improve the balance of risk and return for participants in defined-contribution pension plans, including through increased use of low-cost diversified index funds.

I am testifying before the House Committee on Education on Labor this afternoon on the effects of recent financial market developments on pensions; click here for the testimony.

The key points of the testimony include:

• The turmoil in financial markets has affected many aspects of the economy, including pensions. The most direct effect on pensions is through the prices of financial assets such as corporate equities and bonds. The Standard & Poor’s 500 stock market index, for example, has fallen by more than 25 percent over the past year as the outlook for the economy and corporate profits has worsened. Because the majority of pension assets are held in equities, drops in stock prices have had a significant adverse effect on pension plans.

• Data from the Federal Reserve suggest that the decline in the value of financial assets cost pension funds roughly $1 trillion—almost 10 percent of their assets—from the second quarter of 2007 to the second quarter of 2008, the latest period for which data are available. There has been a significant further drop in asset prices since the end of the second quarter, and it is plausible that the cumulative decline in pension assets over the past year and a half amounts to about $2 trillion.

• In a defined-benefit pension plan, benefits are specified by a fixed formula unrelated to the value of the pension fund. The sponsor of the plan is generally responsible for financing the benefits and must therefore make larger contributions when the value of the assets held by the pension fund declines. By CBO’s estimates, the value of the assets held by defined-benefit plans has declined by roughly 15 percent over the past year. Because of the way the obligations of the plans are calculated, their funding position (that is, the relationship between their assets and liabilities) is also affected by the level of interest rates. Those rates have increased over the past year, lowering the discounted value of plans’ liabilities by roughly 5 percent to 10 percent and partially offsetting the drop in asset values. Overall, according to CBO’s estimates, defined-benefit plans’ assets net of liabilities may have decreased by 5 percent to 10 percent over the past year.

• Changes in asset prices have also affected the value of assets in defined-contribution pension plans. In those plans, the resources available to workers upon retirement depend directly on the value of assets in their plan account. Defined-contribution plans apparently are more heavily weighted toward stocks than defined-benefit plans are; over two-thirds of the assets in defined-contribution plans are invested in equities (either directly or through mutual funds). Because of that heavy emphasis on equities, the value of assets in defined-contribution plans may have declined by slightly more than that of assets in defined-benefit plans.

• Some people on the verge of retirement might respond to a decline in financial markets by working longer. In 2006, 36 percent of people age 65 and older were in families with earnings; that share could rise somewhat over the next few years, both because of underlying trends in the labor market and because of the recent turmoil in financial markets.

• Although severe stresses in financial markets almost inevitably cause wrenching adjustments by workers and employers, the risks can be attenuated by sensibly designing pension plans. For example, although workers enrolled in defined-contribution plans may not be able to avoid bearing the risks associated with broad price changes in financial markets, they can avoid unnecessary risks associated with a lack of diversification. Such unnecessary risks can arise, for example, by overweighting portfolios with individual stocks rather than diversified index funds.

CBO released a brief today on climate-change policy and CO2 emissions from passenger vehicles (for the PDF, click here).

Discussions about addressing climate change (e.g., through a cap-and-trade program or a carbon tax) often focus on the transportation sector. The brief argues, however, that most of the reduction in CO2 emissions would occur in other sectors (e.g., the electricity sector) and that the effects on vehicle emissions would be modest, especially in the shorter run.

To be sure, a cap-and-trade system or a carbon tax would raise the price of gasoline, encouraging consumers to drive less and to buy more fuel-efficient cars– but the magnitude of these effects would be relatively small. For example, CBO has estimated that a price of $28 per metric ton of CO2 in 2012 would lead to a reduction of about 10 percent in total U.S. emissions compared with a no-action scenario. Vehicle emissions, though, would remain relatively constant in the short run, and even over time they would decline only by around 2.5 percent — much less than the 10 percent reduction in overall emissions.

Several factors account for the relatively small influence that a price on CO2 emissions would have on passenger vehicles and driving behavior. First, a CO2 price of $28 per metric ton would raise gas prices by about 25 cents per gallon, far less of an increase than consumers have recently born with little behavioral result. (Between 2003 and 2007, gas prices increased from $1.50 to more than $3.00 per gallon. Vehicle miles driven, driving speeds, and the purchase of larger vehicles have all responded only modestly despite the dramatic increase in prices.) An increase in gas prices of 25 cents or so per gallon is unlikely to generate massive changes in driving behavior.

In addition, recent changes to corporate average fuel economy (CAFE) standards will require substantial gains in fuel economy over the next dozen years. Especially over the longer term, gas price increases are not likely to have a large effect beyond what CAFE standards will require.

Finally, cultural, historic, and geographic considerations drive the extent to which Americans have become dependent on automobile travel, and their choices tend towards larger and more powerful (and less fuel efficient) automobiles. While dramatic increases in gasoline prices (or shifts in cultural norms) might eventually influence these considerations, the magnitude of gas price increases under most legislation under consideration would likely have little effect.

The brief was written by David Austin of our Microeconomic Studies Division.

CBO has just issued its analysis of the financial rescue legislation released by the Senate today. This legislation includes the rescue package considered by the house earlier this week, a temporary increase in deposit insurance, and certain tax provisions including an extension of Alternative Minimum Tax (AMT) relief. The pdf of our analysis is posted here. The text of our analysis is pasted below.

October 1, 2008

Honorable Christopher J. Dodd
Chairman
Committee on Banking, Housing, & Urban Affairs
United States Senate
Washington, DC  20510

Dear Mr. Chairman:

The Congressional Budget Office (CBO) has reviewed the financial rescue legislation to be considered by the Senate.  That legislation contains three separate parts; the bill refers to these three components as “divisions.”

Division A is the Emergency Economic Stabilization Act of 2008, most of which is identical to the financial rescue bill considered by the House of Representatives earlier this week.  In addition to creating a Troubled Asset Relief Program (TARP), under which the Secretary of the Treasury would be authorized to purchase, insure, hold, and sell a wide variety of financial instruments, particularly those that are based on or related to residential or commercial mortgages issued prior to March 14, 2008, Division A also includes a provision that would provide for a temporary increase in federal deposit insurance coverage.

Division B is entitled the Energy Improvement and Extension Act of 2008. It contains numerous tax provisions related to energy production, transportation, and energy conservation.  Division C extends various expiring tax provisions, including alternative minimum tax relief.

Although significant uncertainty surrounds the precise net budgetary impact from Division A of the bill, CBO expects that the bill as a whole (including Divisions B and C) would increase the budget deficit over the next decade.

Division A – Emergency Economic Stabilization Act of 2008

In addition to the TARP, the Senate legislation includes an expansion in deposit insurance.   This section describes and analyzes the provisions of Division A, beginning with the changes to deposit insurance.

Deposit Insurance

Section 136 would provide for a temporary increase in the amount of deposits insured by the Federal Deposit Insurance Corporation (FDIC) and the National Credit Union Administration (NCUA), raising the limit for each insured account from $100,000 to $250,000 through December 31, 2009. Both agencies would be authorized to borrow such sums as may be necessary to cover any additional costs incurred as a result of the expanded coverage.  The legislation also directs the agencies to exclude the increase in insurance coverage when assessing insurance premiums in the near term.

CBO estimates that the deposit insurance funds would incur larger losses in the near term as a result of higher coverage levels and the associated increase in insured deposits.  (When institutions fail, the FDIC and NCUA pay for covered deposits and liquidate the assets held by the institution.  Raising the amount of insured deposits would increase payments to depositors without affecting recoveries from liquidating assets, thereby increasing the net loss to the funds.)  Such near-term losses would, however, be offset over the long term by higher insurance premiums because the agencies are required by law to restore the deposit insurance funds to certain levels over time, so any additional losses from the temporary expansion in coverage will gradually be offset by higher future premiums.

The effects of this provision on outlays over the next year or two are difficult to predict precisely because of uncertainty about the volume and distribution of insured deposits that would be added by this bill. Based on preliminary information from the FDIC, however, CBO estimates that raising the limit to $250,000 through 2009 would boost insured deposits nationwide by about 15 percent.  (As of June 30, 2008, deposits at FDIC-insured institutions totaled about $7 trillion, of which $2.6 trillion were uninsured.  The FDIC estimate suggests that this provision would extend coverage to about $700 billion of those uninsured deposits.)

Overview of TARP

The bill would appropriate such sums as are necessary, for as many years as necessary, to enable the Secretary to purchase or insure troubled assets and to cover all administrative expenses of purchasing, insuring, holding, and selling those assets.  Under the legislation, the authority to enter into agreements to purchase such troubled assets would initially be set to expire on December 31, 2009, but could be extended through two years from the date of enactment upon certification by the Secretary that such an extension is necessary.  The purchase price of all such assets outstanding at any one time could not exceed $700 billion (though cumulative gross purchases could exceed $700 billion as previously purchased assets are sold).  Purchases would be limited as follows:

-Authority for purchases of $250 billion in assets would be available upon enactment;

-The authority would increase to $350 billion if the President submits to the Congress a written notification that the Secretary is exercising authority to purchase an additional $100 billion of assets; and

-The authority would increase to $700 billion if the President submits a report detailing a plan to use the remaining $350 billion in purchase authority; that expansion would be subject to a 15-day Congressional review for potential disapproval of the plan.

The bill would also enable the federal government, under terms and conditions to be developed by the Secretary of the Treasury, to insure troubled assets, including mortgage-backed securities, and collect premiums from participating financial institutions.  The $700 billion limit would be reduced by the excess of obligations to net premiums, if any, under this insurance program.

To facilitate these activities, the federal debt limit would be increased by $700 billion. If, five years after enactment of the bill, the Director of the Office of Management and Budget in consultation with the Director of the Congressional Budget Office determines that the TARP has incurred a net loss, the President would be required to submit a legislative proposal to recoup that shortfall from entities benefiting from the TARP.

Cost of Division A

Under the TARP, the Secretary would have the authority—if deemed necessary to promote stability in the financial markets—to purchase any financial asset at any price and to sell that asset for any price at any future date. That lack of specificity regarding how the authority would be implemented and even what types of assets would be purchased makes it impossible at this point to provide a meaningful estimate of the ultimate impact on the federal budget from enacting this legislation. Although it is not currently possible to quantify the net budget impact given the lack of details about how the program would be implemented, CBO has concluded that enacting Division A would likely entail some net budget cost—which would, however, be substantially smaller than $700 billion. The net budget cost would reflect several factors:

Net gains or losses on the TARP transactions. As noted in CBO’s recent testimony before the House Budget Committee, the net gain or loss on the TARP transactions would reflect the degree to which the federal government sought to obtain, and succeeded in receiving, a fair market price for the assets it purchased, and the degree to which, because of severe market turmoil, market prices would be lower than the underlying value of the assets.

Although some classes of assets and purchase mechanisms are conducive to determining a fair market price, it is unlikely that the program would be limited exclusively to those classes of assets and purchase mechanisms. The program would probably include assets that have the worst credit risks and hence are difficult to price, making it likely that the government would, in some cases, pay prices that fail to cover those risks. Although it is possible that future increases in asset values would generate gains even on assets for which the government initially overpays, an overall net loss is more likely if the government initially overpays.

The bill includes a provision intended to protect against such future net losses by requiring that firms selling troubled assets to the government also provide warrants or senior debt instruments.  CBO anticipates that this provision would not have a substantial effect on the net cost of the TARP, however.  On the one hand, warrants or senior debt instruments might reduce the incentive for sellers to overcharge for low-quality assets.  On the other hand, since the warrants or debt instruments would have value, Treasury would generally face higher prices because sellers would seek compensation for both the value of the troubled asset and the value of the warrant or debt instrument.  In addition, the warrants or senior debt instruments may be difficult for the government to value, complicating even those auctions in which the government is otherwise most likely to obtain a fair market price.

In any case, the ultimate cost to the government on the transactions would not be the total amount spent to purchase assets—limited to $700 billion outstanding at any one time—but rather the difference between the amount spent by the government and the amount received in earnings and sales proceeds when all of the assets are finally sold, presumably some years from now. That net cost is likely to be substantially less than $700 billion but is more likely than not to be greater than zero.

Recoupment mechanism. The recoupment mechanism is designed to offset any net losses the government experiences on the TARP transactions. The mechanism, however, requires only that the President submit a proposal to offset such costs after five years. Even if it would be fully effective in offsetting any net losses, the President’s proposal would require a future act of Congress to be implemented. Any savings from such legislation would be estimated when the proposal is considered and would be credited to that legislation for Congressional scorekeeping purposes.

Administrative costs. Beyond the effect of any gains or losses on the transactions under the TARP and the recoupment mechanism, the programs authorized by this bill would involve administrative costs. For example, the government would have to compensate the private asset managers hired by the Treasury. Those administrative costs are not included in the $700 billion limit on asset purchases. Even if the transactions and the recoupment mechanism combined resulted in neither a gain nor a loss for the government, the administrative costs would expand the budget deficit.

The legislation includes a variety of other provisions that would, on net, add to the budget deficit. A number of those provisions are discussed below.

Other Major Provisions of Division A

In addition to the expansion in FDIC insurance limits and the provisions of the TARP discussed above, Division A also contains provisions that would:

-Change the tax treatment of certain types of income, losses, or deductions of corporations or individuals;

-Require that certain financial institutions seeking to sell assets through the TARP meet appropriate standards for senior executive officers’ compensation, as determined by the Secretary of the Treasury;

-Require the Secretary of the Treasury to take steps to maximize assistance for homeowners, including encouraging servicers of the underlying mortgages to take advantage of the Hope for Homeowners Program under section 257 of the National Housing Act;

-Allow the Federal Reserve System to pay interest on certain reserves of depository institutions that are held on deposit at the Federal Reserve, starting on October 1, 2008;

-Direct the Federal Housing Finance Agency, the Federal Deposit Insurance Corporation, and the Federal Reserve Board to implement various measures with regard to residential loans and securities under their control in order to reduce the number of foreclosures, which could include modifying the terms of such loans; and

-Establish Congressional oversight and reporting requirements related to implementation of the legislation, along with a Financial Stability Oversight Board with responsibility for overseeing operations of the program.

The bill would require that the federal budget display the costs of purchasing or insuring troubled assets using procedures similar to those specified in the Federal Credit Reform Act, but adjusting for market risk (in a manner not reflected in that law). In particular, the federal budget would not record the gross cash disbursements for purchases of troubled assets (or cash receipts for their eventual sale), but instead would reflect the estimated net cost to the government of such purchases (broadly speaking, the purchase cost minus the present value, adjusted for market risk, of any estimated future earnings from holding those assets and the proceeds from the eventual sale of them).

Impact on Federal Finances

CBO expects that the Treasury would use most or all of the $700 billion in purchase authority within two years (after which the authority to enter into agreements to purchase various troubled assets would expire). To finance those purchases, the Treasury would have to sell debt to the public. Federal debt held by the public would therefore rise by about $700 billion, although the government would also acquire valuable financial assets in the process.  As noted above, CBO expects that since the acquired assets would have some value, the net budget impact would be substantially less than $700 billion; similarly, net cash disbursements under the program would also be substantially less than $700 billion over time because, ultimately, the government would sell the acquired assets and thus generate income that would offset much of the initial expenditures.

In addition to any net gain or loss on the purchase of $700 billion or more in assets, the government also would incur administrative costs for the proposed program. Those costs would depend on the kinds of assets purchased or insured. On the basis of the costs incurred by private investment firms that acquire, manage, and sell similar assets, CBO expects that the administrative costs of operating the program could amount to a few billion dollars per year, as long as the government held all or most of the purchased assets.

Other provisions in Division A would on net increase the budget deficit.  For example, the legislation would allow the Federal Reserve to pay interest immediately on certain reserve balances of depository institutions, rather than starting on October 1, 2011, as allowed under current law.  CBO estimates that, over the next three years, the provision would reduce the Federal Reserve’s payments of its profits to the Treasury, which are classified as revenue in the federal budget.

In addition, a number of provisions Division A would affect federal revenues by changing tax law, including provisions that would limit the deductibility of executive compensation for certain firms selling assets; allow losses incurred by certain taxpayers on preferred stock in Fannie Mae and Freddie Mac to be treated as ordinary rather than capital losses; and exclude from income amounts attributable to the cancellation of mortgage debt of individuals in certain circumstances. The Joint Committee on Taxation estimates that, on net, these provisions in Division A would reduce federal revenues.

Enacting Division A could also affect other federal spending—including, for example, outlays from the operations of Fannie Mae, Freddie Mac, federal housing programs, and deposit insurance. Some of those effects would be related to how TARP would be used to purchase assets (including what kinds of assets would be acquired and from what types of institutions), and how successful the program would be in restoring liquidity to the nation’s financial markets.

Division B – Energy Improvement and Extension Act of 2008

Division B of the bill would provide a number of tax incentives related to energy and fuel production and energy conservation.  It also includes several provisions that would raise revenue, with the largest effect from a modification of the requirements imposed on brokers for the reporting of their customers’ basis in securities transactions.  CBO and the Joint Committee on Taxation (JCT) estimate that, over the 2009-2013 period, Division B would reduce revenues by $6.8 billion, increase outlays by about $0.2 billion, and increase projected deficits by about $7 billion.  CBO and JCT estimate that, over the 2009-2018 period, Division B would increase revenues by about $0.3 billion, increase outlays by about $0.2 billion, and reduce projected deficits by less than $0.1 billion.

Division C – Tax Extensions and Alternative Minimum Tax Relief

Division C would extend relief from the alternative minimum tax for 2008, extend and modify a number of other expiring tax provisions, provide tax relief for regions of the country affected by severe storms earlier this year, make other changes to tax law, and provide payments to state and local governments for support to rural schools and other county programs.  It also would modify the tax treatment of deferred compensation paid by certain foreign entities.  CBO and JCT estimate that, over the 2009-2013 period, Division C would reduce revenues by about $105.2 billion, increase outlays by $7.1 billion, and increase projected deficits by about $112.3 billion.  CBO and JCT estimate that, over the 2009-2018 period, Division C would reduce revenues by about $99.5 billion, increase outlays by about $7.5 billion, and increase projected deficits by about $107.1 billion.

Intergovernmental and Private-Sector Mandates

The non-tax provisions of the legislation contain no intergovernmental mandates as defined in the Unfunded Mandates Reform Act (UMRA).

The non-tax provisions do, however, contain private-sector mandates as defined in UMRA, and CBO estimates that the aggregate cost of those mandates would exceed the annual threshold established in UMRA ($136 million in 2008, adjusted annually for inflation).

Division C, section 512 would impose a private-sector mandate on group health plans and issuers of group health insurance by prohibiting them from imposing treatment limitations or financial requirements for mental health benefits that differ from those placed on medical and surgical benefits. CBO estimates that the direct costs of the private-sector mandate would significantly exceed the annual threshold established in UMRA in each of the first five years that the mandate would be in effect.

Division A, section 136 could impose a private-sector mandate to the extent that deposit insurance premiums are higher than they would be in the absence of  this bill. Most depository institutions (commercial banks, savings associations, and credit unions) are required by law to have federal deposit insurance. CBO, therefore, considers changes in the federal deposit insurance system that increase requirements on those institutions to be private-sector mandates under UMRA. The cost of the mandate would be the additional premiums assessed during each of the first five years the mandate is in effect. While CBO expects that any additional losses from the temporary expansion in coverage would gradually be offset by higher future premiums, we cannot estimate the cost of the mandate because of uncertainty about the timing of the losses and whether or by how much premiums would increase during those first five years.

I hope this information is helpful to you. If you have further questions about CBO’s analysis, do not hesitate to contact me.

Sincerely,

Peter R. Orszag
Director

cc:

Honorable Richard C. Shelby
Ranking Member

Honorable Kent Conrad
Chairman
Committee on the Budget

Honorable Judd Gregg
Ranking Member

Honorable Max Baucus
Chairman
Committee on Finance

Honorable Charles E. Grassley
Ranking Member

CBO has just issued its analysis of the Emergency Economic Stabilization Act of 2008, as released tonight by the House Committee on Financial Services. Among other provisions, the legislation would create a Troubled Asset Relief Program (TARP). The pdf of our analysis is posted here. The text is pasted below.

—

September 28, 2008

Honorable Barney Frank
Chairman
Committee on Financial Services
U.S. House of Representatives
Washington, DC 20515

Dear Mr. Chairman:

The Congressional Budget Office (CBO) has reviewed the Emergency Economic Stabilization Act of 2008, as released by the House Committee on Financial Services on September 28, 2008. The legislation would, among other provisions, create a Troubled Asset Relief Program (TARP), under which the Secretary of the Treasury would be authorized to purchase, insure, hold, and sell a wide variety of financial instruments, particularly those that are based on or related to residential or commercial mortgages issued prior to March 14, 2008. Under the legislation, the authority to enter into agreements to purchase such troubled assets would initially be set to expire on December 31, 2009, but could be extended through two years from the date of enactment upon certification by the Secretary that such an extension is necessary.

The bill would appropriate such sums as are necessary, for as many years as necessary, to enable the Secretary to purchase or insure troubled assets and to cover all administrative expenses of purchasing, insuring, holding, and selling those assets. The purchase price of all such assets outstanding at any one time could not exceed $700 billion (though cumulative gross purchases could exceed $700 billion as previously purchased assets are sold). Purchases would be limited as follows:

- Authority for purchases of $250 billion in assets would be available upon enactment;

- The authority would increase to $350 billion if the President submits to the Congress a written notification that the Secretary is exercising authority to purchase an additional $100 billion of assets; and

- The authority would increase to $700 billion if the President submits a report detailing a plan to use the remaining $350 billion in purchase authority; that expansion would be subject to a 15-day Congressional review for potential disapproval of the plan.

The bill would also enable the federal government, under terms and conditions to be developed by the Secretary of the Treasury, to insure troubled assets, including mortgage-backed securities, and collect premiums from participating financial institutions. The $700 billion limit would be reduced by the excess of obligations to net premiums, if any, under this insurance program.

To facilitate these activities, the federal debt limit would be increased by $700 billion. If, five years after enactment of the bill, the Director of the Office of Management and Budget in consultation with the Director of the Congressional Budget Office determines that the TARP has incurred a net loss, the President would be required to submit a legislative proposal to recoup that shortfall from entities benefiting from the TARP.

Cost of the Legislation

Under the TARP, the Secretary would have the authority—if deemed necessary to promote stability in the financial markets—to purchase any financial asset at any price and to sell that asset for any price at any future date. That lack of specificity regarding how the authority would be implemented and even what types of assets would be purchased makes it impossible at this point to provide a meaningful estimate of the ultimate impact on the federal budget from enacting this legislation. Although it is not currently possible to quantify the net budget impact given the lack of details about how the program would be implemented, CBO has concluded that enacting the bill would likely entail some net budget cost—which would, however, be substantially smaller than $700 billion. The net budget cost would reflect several factors:

Net gains or losses on the TARP transactions. As noted in CBO’s recent testimony before the House Budget Committee, the net gain or loss on the TARP transactions would reflect the degree to which the federal government sought to obtain, and succeeded in receiving, a fair market price for the assets it purchased, and the degree to which, because of severe market turmoil, market prices would be lower than the underlying value of the assets.

Although some classes of assets and purchase mechanisms are conducive to determining a fair market price, it is unlikely that the program would be limited exclusively to those classes of assets and purchase mechanisms. The program would probably include assets that have the worst credit risks and hence are difficult to price, making it likely that the government would, in some cases, pay prices that fail to cover those risks. Although it is possible that future increases in asset values would generate gains even on assets for which the government initially overpays, an overall net loss is more likely if the government initially overpays.

The bill includes a provision intended to protect against such future net losses by requiring that firms selling troubled assets to the government also provide warrants or senior debt instruments. CBO anticipates that this provision would not have a substantial effect on the net cost of the TARP, however. On the one hand, warrants or senior debt instruments might reduce the incentive for sellers to overcharge for low-quality assets. On the other hand, since the warrants or debt instruments would have value, Treasury would generally face higher prices because sellers would seek compensation for both the value of the troubled asset and the value of the warrant or debt instrument. In addition, the warrants or senior debt instruments may be difficult for the government to value, complicating even those auctions in which the government is otherwise most likely to obtain a fair market price.

In any case, the ultimate cost to the government on the transactions would not be the total amount spent to purchase assets—limited to $700 billion outstanding at any one time—but rather the difference between the amount spent by the government and the amount received in earnings and sales proceeds when all of the assets are finally sold, presumably some years from now. That net cost is likely to be substantially less than $700 billion but is more likely than not to be greater than zero.

Recoupment mechanism. The recoupment mechanism is designed to offset any net losses the government experiences on the TARP transactions. The mechanism, however, requires only that the President submit a proposal to offset such costs after five years. Even if it would be fully effective in offsetting any net losses, the President’s proposal would require a future act of Congress to be implemented. Any savings from such legislation would be estimated when the proposal is considered and would be credited to that legislation for Congressional scorekeeping purposes.

Administrative costs. Beyond the effect of any gains or losses on the transactions under the TARP and the recoupment mechanism, the programs authorized by this bill would involve administrative costs. For example, the government would have to compensate the private asset managers hired by the Treasury. Those administrative costs are not included in the $700 billion limit on asset purchases. Even if the transactions and the recoupment mechanism combined resulted in neither a gain nor a loss for the government, the administrative costs would expand the budget deficit.

The legislation includes a variety of other provisions that would, on net, add to the budget deficit. A number of those provisions are discussed below.

Other Major Provisions

The bill also contains provisions that would:

- Change the tax treatment of certain types of income, losses, or deductions of corporations or individuals;

- Require that certain financial institutions seeking to sell assets through the TARP meet appropriate standards for senior executive officers’ compensation, as determined by the Secretary of the Treasury;

- Require the Secretary of the Treasury to take steps to maximize assistance for homeowners, including encouraging servicers of the underlying mortgages to take advantage of the Hope for Homeowners Program under section 257 of the National Housing Act;

- Allow the Federal Reserve System to pay interest on certain reserves of depository institutions that are held on deposit at the Federal Reserve, starting on October 1, 2008;

- Direct the Federal Housing Finance Agency, the Federal Deposit Insurance Corporation, and the Federal Reserve Board to implement various measures with regard to residential loans and securities under their control in order to reduce the number of foreclosures, which could include modifying the terms of such loans; and

- Establish Congressional oversight and reporting requirements related to implementation of the legislation, along with a Financial Stability Oversight Board with responsibility for overseeing operations of the program.

The bill would require that the federal budget display the costs of purchasing or insuring troubled assets using procedures similar to those specified in the Federal Credit Reform Act, but adjusting for market risk (in a manner not reflected in that law). In particular, the federal budget would not record the gross cash disbursements for purchases of troubled assets (or cash receipts for their eventual sale), but instead would reflect the estimated net cost to the government of such purchases (broadly speaking, the purchase cost minus the present value, adjusted for market risk, of any estimated future earnings from holding those assets and the proceeds from the eventual sale of them).

Impact on Federal Finances

CBO expects that the Treasury would use most or all of the $700 billion in purchase authority within two years (after which the authority to enter into agreements to purchase various troubled assets would expire). To finance those purchases, the Treasury would have to sell debt to the public. Federal debt held by the public would therefore rise by about $700 billion, although the government would also acquire valuable financial assets in the process. As noted above, CBO expects that since the acquired assets would have some value, the net budget impact would be substantially less than $700 billion; similarly, net cash disbursements under the program would also be substantially less than $700 billion over time because, ultimately, the government would sell the acquired assets and thus generate income that would offset much of the initial expenditures.

In addition to any net gain or loss on the purchase of $700 billion or more in assets, the government also would incur administrative costs for the proposed program. Those costs would depend on the kinds of assets purchased or insured. On the basis of the costs incurred by private investment firms that acquire, manage, and sell similar assets, CBO expects that the administrative costs of operating the program could amount to a few billion dollars per year, as long as the government held all or most of the purchased assets.

Other provisions in the legislation would on net increase the budget deficit. For example, the legislation would allow the Federal Reserve to pay interest immediately on certain reserve balances of depository institutions, rather than starting on October 1, 2011, as allowed under current law. CBO estimates that, over the next three years, the provision would reduce the Federal Reserve’s payments of its profits to the Treasury, which are classified as revenue in the federal budget.

In addition, a number of provisions in the bill would affect federal revenues by changing tax law, including provisions that would limit the deductibility of executive compensation for certain firms selling assets; allow losses incurred by certain taxpayers on preferred stock in Fannie Mae and Freddie Mac to be treated as ordinary rather than capital losses; and exclude from income amounts attributable to the cancellation of mortgage debt of individuals in certain circumstances. The Joint Committee on Taxation estimates that, on net, these provisions would reduce federal revenues.

Enacting the legislation could also affect other federal spending—including, for example, outlays from the operations of Fannie Mae, Freddie Mac, federal housing programs, and deposit insurance. Some of those effects would be related to how TARP would be used to purchase assets (including what kinds of assets would be acquired and from what types of institutions), and how successful the program would be in restoring liquidity to the nation’s financial markets.

Intergovernmental and Private-Sector Mandates

The non-tax provisions of the bill would impose no intergovernmental or private-sector mandates as defined in the Unfunded Mandates Reform Act.

I hope this information is helpful to you. If you have further questions about CBO’s analysis, do not hesitate to contact me.

Sincerely,
Peter R. Orszag
Director

cc:

Honorable Spencer Bachus
Ranking Member

Honorable John M. Spratt Jr.
Chairman
Committee on the Budget

Honorable Paul Ryan
Ranking Member

Identical letter sent to the Honorable Christopher J. Dodd.

CBO has released a cost estimate for HR 7174, the James Zadroga 9/11 Health and Compensation Act of 2008.

The legislation would provide:

• Health care benefits for eligible emergency personnel who responded to the terrorist attacks in New York City on September 11, 2001, and for recovery and clean-up workers following the attacks;

• Health care benefits to eligible residents and others present in the part of New York City that was affected by those attacks; and

• Monetary compensation to newly eligible individuals for death and physical injury claims resulting from the attacks.

In addition, the legislation would raise revenues by altering various provisions of the tax code.

CBO estimates that enacting H.R. 7174 would increase direct spending by just under $11 billion over the 2009-2018 period. The Joint Committee on Taxation (JCT) estimates that the tax provisions in the bill would increase revenues by about $11 billion over the same period. On balance, CBO and JCT estimate that the direct spending and revenue effects from enacting the legislation would reduce deficits by about $230 million over the 2009-2013 period and by $35 million over the 2009-2018 period.

A Wall Street Journal blog posting mischaracterizes CBO’s testimony earlier this week on the net budget impact of the Treasury proposal to buy troubled assets. The Wall Street Journal blog states that the plan “likely won’t have any effect on the 2009 budget deficit.” That is incorrect.

Here’s what I said in the testimony:

“In particular, the federal budget would not record the gross cash outlays associated with purchases of troubled assets but, instead, would reflect the estimated net cost to the government of such purchases (broadly speaking, the purchase cost minus the expected value of any estimated future earnings from holding those assets and the proceeds from the eventual sale of them). In CBO’s view, that budgetary treatment best reflects the impact of the purchases of financial assets on the federal government’s underlying financial condition. The fundamental idea is that if the government buys a security at the going market price, it has exchanged cash for another asset rather than caused a deterioration in its underlying fiscal position.”

The testimony then noted that even though the gross cash outlays would perhaps amount to $700 billion, the net budget impact would be substantially smaller because the government would be acquiring assets with some value. It did not say, though, that the net budget impact would be zero, as the Wall Street Journal suggests.

So what will the net budget impact be, even if it’s substantially smaller than $700 billion? That depends on three factors: (a) the degree to which the transactions result in a gain or loss to the government; (b) the administrative costs of running the program; and (c) any interactive effects with other government programs. The first issue is central, and the testimony noted that “whether those transactions ultimately resulted in a gain or loss to the government would depend on the types of assets purchased, how they were acquired and managed, and when and under what terms they were sold.”

The testimony went on to explore the forces that could affect any gain or loss. “In addition to the future evolution of the housing prices, interest rates, and other fundamental drivers of asset values, two key forces would influence the net gain or loss on the assets purchased:

• Whether the federal government seeks and is able to succeed in obtaining a fair market price for the assets it purchases and, in particular, whether it can avoid being saddled with the worst credit risks without the purchase price reflecting those risks. Concerns about the government’s overpaying are particularly salient when sellers offer assets with varying underlying characteristics that are complicated to evaluate.  Such problems are attenuated the more that the government focuses on buying part of a given asset from institutions that all own a share of that asset, rather than buying different assets from different institutions. That is, the government is more likely to pay a fair price when multiple institutions are competing to sell identical assets than when it has to assess competing offers for different assets with hard-to-determine values.

• Whether, because of severe market turmoil, market prices are currently lower than the underlying value of the assets. If current prices reflect “fire sale” prices that can result from severe liquidity constraints and the impairment of credit flows, then taxpayers could possibly benefit along with the institutions selling the assets. Under normal circumstances, prices do not long depart from their fundamentals because the incentive to engage in arbitrage and profit from price discrepancies is large. But arbitrage practices work less well when liquidity is restrained, as it is now, and many potential arbitragers cannot get short-term financing. It is therefore at least possible that the prices of some assets are below their fundamental value; in that case, to the extent that the government bought now and held such assets until their market prices recovered to reflect that underlying value, net gains would be possible.”

Nothing in CBO’s testimony should be interpreted as suggesting that the interplay between these two forces would generate a net impact of zero for the transactions alone. Indeed, although the lack of specificity in the bill means that CBO cannot currently quantify its net budgetary impact, and although there is some possibility that the government could realize a net gain on the transactions authorized under the bill, it seems more likely that enacting the bill would result in an increase in the federal deficit. In other words, the net budgetary cost (including administrative costs) is very likely to be substantially smaller than $700 billion, but it seems likely to be greater than zero.