2008 financial rescue and TARP (United States, 2008 to 2023)

In October 2008, after the House first voted the plan down, Congress passed a law letting Treasury spend up to $700 billion to stabilize the financial system, and the final official accounting (CBO, 2024) put the net cost at about $31 billion, mostly from housing programs, while critics and defenders still disagree about whether it worked and whether the public was told enough.

The record

This case is about how a very large, fast spending decision was made, how it was overseen, and what the public could learn. It does not decide whether the rescue was right. The numbers below are limited to what official bodies state, with the source named for each.

What we could and could not open. Primary sources we opened: the Congressional Budget Office (CBO) final report page (April 2024); the final SIGTARP quarterly letter (Q1 fiscal 2024); the Congressional Oversight Panel (COP) final press release (March 16, 2011, on the St. Louis Fed FRASER archive); the testimony of Special Inspector General Neil Barofsky (July 2009) on the House Financial Services Committee site; 12 U.S.C. 5201 (purposes) and 5233 (COP) on Cornell's Legal Information Institute; and the House Oversight Committee release of July 21, 2009. Could not open: the Treasury TARP page (returned only metadata), the text of Public Law 110-343 on congress.gov (blocked by robots rules), Ben Bernanke's Sept. 23, 2008 testimony (fetch error), and the full COP final report. Roll call counts for the three votes come from Wikipedia and were not checked against the House Clerk or Senate.gov.

Timeline

• Sept. 2008: Treasury Secretary Henry Paulson sends Congress a draft authorizing the purchase of troubled assets. Wikipedia says the draft was three pages and included a section making Treasury decisions non-reviewable; the final law dropped it and made Treasury actions subject to judicial review (secondary; the draft itself was not opened).
• Sept. 19 to 24, 2008: Polls reported by Wikipedia: Pew, 57 percent supported government action and 30 percent opposed; a Bloomberg/LA Times poll found 55 percent opposed using taxpayer dollars to rescue private firms; a USA Today/Gallup poll (Sept. 24) found 56 percent wanted Congress to pass a different plan, 22 percent backed Paulson's plan in its initial form and 11 percent wanted no action (secondary; polls not opened).
• Sept. 29, 2008: The House rejects the bill 205 to 228 (Wikipedia).
• Oct. 1, 2008: The Senate passes a revised bill 74 to 25 (Wikipedia).
• Oct. 3, 2008: The House passes it 263 to 171, and President George W. Bush signs the Emergency Economic Stabilization Act of 2008 (Public Law 110-343) the same day (Wikipedia). Wikipedia reports that 33 Democrats and 24 Republicans who had voted no on Sept. 29 voted yes.
• Funding structure: $700 billion authorized. The Treasury Secretary had access to the first $250 billion, the President could release another $100 billion, and the last $350 billion was subject to a 15-day congressional disapproval window (Wikipedia).
• Oversight structure: The Act created a Congressional Oversight Panel (five members appointed by congressional leaders, "an establishment in the legislative branch," 12 U.S.C. 5233, opened) and a Special Inspector General for TARP (SIGTARP), among other bodies. The stated purposes include "to immediately provide authority and facilities" to stabilize the system, and using it in a manner that protects home values, college funds, retirement accounts and life savings (12 U.S.C. 5201, opened).
• Jan. to Feb. 2009: CBO estimated a $64 billion subsidy cost for the first $247 billion; the COP found Treasury paid $254 billion for assets worth about $176 billion (both via Wikipedia's summary, secondary; we did not open these reports).
• July 2009: Barofsky testifies that Treasury "repeatedly" rejected transparency recommendations; the House Oversight Committee holds a hearing on Treasury's lack of TARP transparency (details below).
• Mar. 16, 2011: COP releases its 30th and final report; it terminated on April 3, 2011 by statute (FRASER press release, opened).
• Dec. 2014: Wikipedia, citing Treasury via the Wall Street Journal, reports Treasury sold its remaining Ally Financial stake on Dec. 19, 2014, with $441.7 billion recovered on $426.4 billion invested (secondary).
• Apr. 28, 2023: The Home Affordable Modification Program (HAMP) ended (SIGTARP final letter, opened).
• May 31, 2023: Last $642 million for mortgage programs deobligated. Sept. 2023: last investment repaid, ending the program (CBO, opened).
• Dec. 31, 2023: SIGTARP issues its final quarterly report (opened).
• Apr. 2024: CBO final report on TARP.

The money: three official numbers that look different and are not

The numbers differ because they count different things. CBO's April 2024 report estimates a net gain of $9 billion on capital purchases and other support for financial institutions, and Treasury's own cash-basis table shows $16.3 billion of income on its largest bank program, the Capital Purchase Program. CBO's $31 billion total adds the programs that cost money: mortgage programs ($31 billion), which were mostly grants that were never expected to be repaid, and auto assistance ($12 billion). Treasury's $31.1 billion net cost (as of September 30, 2023) agrees on the total, though the two count differently: CBO estimates subsidy cost, while Treasury reports cash disbursed and collected plus $13.1 billion of interest expense. The two also group programs differently: Treasury lists AIG on its own, at a cost of $15.2 billion, while CBO counts AIG inside financial institutions. Early estimates were much higher: COP's final press release says CBO's estimate fell to $25 billion from the $356 billion first projected. SIGTARP's final letter gives no total cost. Larger headline totals seen in news coverage were not examined and are not used.

Enforcement and oversight record

SIGTARP's final letter (opened) states: recoveries "exceed $11.4 billion"; 476 defendants prosecuted, 328 sentenced to prison, including 75 bankers; enforcement actions against 25 corporations or entities; 121 convicted scammers in HAMP-related cases. The letter is signed by Principal Deputy Special Inspector General Melissa Bruce.

What the public was told, and when

• Use of funds. Barofsky's July 22, 2009 testimony (opened) says Treasury had "repeatedly" rejected SIGTARP recommendations on basic transparency, that taxpayers were not told what most TARP recipients did with the money, and that taxpayers were not told how much their investments were worth. SIGTARP had recommended that all recipients report how they used funds; Treasury declined, calling such reporting "meaningless" because money is fungible, with exceptions for Citigroup, Bank of America and AIG. A SIGTARP survey (July 20, 2009) found most banks did not track TARP money dollar for dollar but could describe use: some said it kept them from shrinking lending, others used it to acquire institutions, invest in securities, or hold as a cushion.
• Oversight Committee. The House Oversight and Government Reform Committee, Chairman Edolphus Towns, held a hearing on Treasury and TARP transparency on July 21, 2009 (committee release, opened). Ranking Member Darrell Issa (R-CA) asked that Treasury Secretary Timothy Geithner be called. Towns said the committee would post data itself if Treasury did not publish fund use, portfolio value and lending reports. The release does not include Treasury's response. (Barofsky's own dated testimony is July 22; the two items may be two hearings. We did not resolve this.)
• COP final report. The COP said "Treasury sent tens of billions of dollars to large institutions early on without requiring them to disclose how the money was used" (the FRASER release's wording, summarized by the fetch tool).
• Federal Reserve disclosure (related, not TARP). The Supreme Court on March 21, 2011 declined to overturn two lower-court rulings requiring the Federal Reserve to disclose details of 2008 crisis aid after Bloomberg and Fox News sued under the Freedom of Information Act (FreedomInfo.org, secondary). This was about Fed programs, not TARP, but it shows how recipient names reached the public through court action.

What this case shows

1. A very large decision was made in about two weeks, and the public voice shows up in the votes. The House first voted no on Sept. 29, 2008 (205 to 228). Within four days, 33 Democrats and 24 Republicans who had voted no switched and it passed 263 to 171 (Wikipedia). Polls in the same month showed no majority for the plan in its first form (Gallup: 56 percent wanted a different plan). Voters could not follow the details in days, but their representatives did respond to pressure, and the bill that passed differed from the draft, including judicial review.
2. Oversight was built into the law, and it worked as a disclosure channel. The COP, SIGTARP, and a Comptroller General audit gave Congress and the public a regular, independent record. The public learned costs, losses and fraud from those bodies, not from Treasury alone: COP's February 2009 finding of a $78 billion shortfall on early purchases, Barofsky's 476 prosecutions, CBO's final accounting. Statutes like this are a concrete answer to "what could ordinary people fix."
3. Where the executive resisted disclosure, the public learned less. Barofsky said Treasury rejected transparency recommendations; COP said early funds went out without disclosure of use. Treasury's stated reason, that money is fungible, is a reasonable point and is recorded below, but the effect was that voters could not see what their money did. The Fed disclosure suit shows that recipient names reached the public by lawsuit, not by choice.
4. The headline number depends on what is counted. A net gain of $9 billion on bank support (CBO) and a net cost of $31 billion for all of TARP (CBO; $31.1 billion by Treasury's count) are both official and both correct for their scopes. A voter who hears only one cannot judge. A disclosure rule that requires agencies to publish both, with definitions, would serve the public.
5. Promised goals and results diverged and the target was never formally revised. COP's final press release says HAMP was meant to prevent 3 to 4 million foreclosures and was on track to help 700,000 to 800,000 homeowners, and that Treasury "never formally revised its target." People who voted on the plan on the understanding that homeowners would be helped could not easily see the gap.
6. The long-run effect on expectations is a public question. COP said TARP reinforced "too big to fail" expectations and created moral hazard (FRASER release); the Fiscal Times reports former Sen. Ted Kaufman, COP chair, called the cost of moral hazard "real and it is reprehensible." This is about whether the public was asked to accept a standing guarantee without a vote on it.

The strongest objections, and our answers

1. Objection: The rescue worked and prevented a second Great Depression. Support: Fed Chairman Ben Bernanke warned of "a cataclysm that could have rivaled or surpassed the Great Depression" (quoted in the COP release); the COP also said TARP gave "critical support" to markets; Wikipedia cites a 2012 economist survey agreeing unemployment would have been higher at the end of 2010 without the program. Mitigation: The COP, which wrote that, also said TARP "doesn't deserve full credit" for averting a second Great Depression. We present the counterfactual as contested, not settled, and do not score the decision.
2. Objection: TARP made money, so the criticism is misplaced. Support: CBO's estimated $9 billion net gain on capital purchases and other support for financial institutions (CBO). Mitigation: CBO's final figure for all of TARP is a $31 billion cost, mainly mortgage programs. Both are on the record. COP also warned that low cost "understates risk" because some non-TARP programs moved costs off TARP's books (COP release). We say which number we mean each time.
3. Objection: The transparency criticism ignores that money is fungible. Support: Treasury called use-of-funds reporting "meaningless" for that reason (Barofsky testimony). Mitigation: Barofsky's July 2009 audit found banks could describe actual or planned uses even if they did not track dollars, and Treasury required reporting from Citigroup, Bank of America and AIG. The disagreement between a named inspector general and Treasury is recorded; we do not decide it.
4. Objection: Critics also include the left and right, so this is not about one side. Support: Fiscal Times quotes J. Mark McWatters (a Republican appointee to COP) that TARP "reinforced the bubble bailout cycle," and Damon Silvers (AFL-CIO, a Democratic appointee) pressing Treasury on HAMP. Defense by Timothy Massad of Treasury that Dodd-Frank gives "a variety of tools." Mitigation: Agreed. We attribute opinions to named people and avoid party scoring.
5. Objection: Congress voted twice on an amended bill; consent was given. Support: the vote record. Mitigation: True. The point is about what was knowable in the days between, not that no vote occurred. The public later learned that the first $250 billion went mostly to large institutions without conditions on use (COP).
6. Objection: The Bloomberg lawsuit concerned the Fed, not TARP. Mitigation: Correct, and we label it as related.
7. Objection: The "TARP" label covers different things, and mixing them is misleading. Mitigation: We separate bank investments, AIG, autos and mortgage programs in the table, using CBO categories.

Open questions: what would change our minds

• Vote counts. The 205 to 228, 74 to 25 and 263 to 171 counts, and the switch of 33 Democrats and 24 Republicans, come from Wikipedia. Check the House Clerk and Senate.gov.
• Text of Public Law 110-343. Not opened. Section numbers for SIGTARP, judicial review and the 15-day disapproval window should be checked against the statute, not Wikipedia.
• The Paulson draft. The three-page length and the non-reviewability section are from Wikipedia. Open the original draft.
• Polling. The Pew, Bloomberg/LA Times and Gallup numbers are from Wikipedia. Check the pollsters.
• Dodd-Frank reduction to $475 billion. Treasury's TARP page (opened) says the authority was reduced to $475 billion by Dodd-Frank. We did not read the statute.
• Early cost figures. The CBO January 2009 estimate and COP February 2009 shortfall are from Wikipedia's summary. The COP release says CBO's later estimate fell to $25 billion from $356 billion first projected; we did not open the CBO documents.
• The two Barofsky hearings. The House Financial Services testimony is dated July 22, 2009, and the Oversight release July 21. Confirm whether these are two hearings.
• Treasury's response to SIGTARP on transparency. Not opened; later disclosures (for example whether Treasury later required use-of-funds reporting) are unverified.
• The earlier $15.3 billion surplus figure. An earlier version of this page gave $15.3 billion (Wikipedia, citing Treasury and the WSJ). Treasury's current page does not show it, so it was removed. We use CBO's $9 billion net gain and Treasury's $16.3 billion Capital Purchase Program income.
• Counterfactual. Whether the rescue prevented a second Depression cannot be settled from these sources. A review of the academic literature (for example, the economist survey Wikipedia cites) was not done.
• Fed programs. Out of scope: Fed lending and the AIG non-TARP funds.
• If the full SIGTARP and COP reports show Treasury later adopted use-of-funds reporting broadly, argument 3 would soften.

Sources

Found an error or a better source? We keep every correction on the record with its date and reason. Nothing here is quietly rewritten.