While Cass is on target in pointing to the waste and abuse in the financial sector, the big question is where is the rest of the economics profession? Supposedly, eliminating waste and corruption was the mantra of “free trade” neo-liberals. But the massive waste and corruption in the financial sector is easy for anyone with clear eyes to see. - Dean Baker
Everyone who reads this publication knows what needs to get built. More grid transmission. More nuclear plants. More housing. More automated industrial capacity. The policy conversation has gotten good at identifying the demand side: permitting reform, regulatory overhaul, federal lending, emergency declarations. All essential, all should continue.
What almost nobody talks about is the supply side of capital. Scratch that, a lot of critics keeps talking about how YIMBYs, industrial policy wonks, and the rest don’t have a clue how to finance their proposals (despite the mountain of literature saying otherwise). That capital exists. There is roughly $6 trillion of it, sitting in American public pension funds, currently being allocated in a way that serves nobody.
Oren Cass’s recent New York Times column describes an American financial sector that has stopped fueling the real economy and started consuming it. Dean Baker, writing at the Center for Economic and Policy Research, endorsed Cass’s diagnosis and added a complementary set of regulatory reform proposals. Both focus on constraining finance: transaction taxes, bankruptcy reform, cultural stigma. Neither addresses where the capital should go instead, or why the largest pool of patient capital in the American economy is the right place to start.
A conservative populist and a progressive labor economist who has called finance a cancer on the economy, arriving at the same diagnosis. I think that’s interesting, especially given how much the capital question matters for everything I care about.
A note on the economics profession’s credibility problem
Part of why people mistrust economics, is the fact that most people see only one side of the profession. The side whose’s public voice applies its skepticism selectively and has an incredibly poor track record both in economics and in finance. Larry Summers warned against “moral hazard lectures” and demanded SVB depositors be made whole immediately in 2023, months after calling student loan relief inflationary and unfair. Moral hazard for borrowers, bailouts for banks. Not lost on the public.
The convergence myth
The countries that followed the more “orthodox” development playbook didn’t converge with rich ones, despite the “advocacy” of that side of the profession. The country that built the world’s largest development bank, directed state capital into infrastructure at a third the cost, and added more industrial robots in 2023 than the rest of the world combined did. China’s state-directed industrial policy accounts for roughly three-quarters of the global decline in extreme poverty over four decades. David Oks noted that the three economists who announced a “new era of unconditional convergence” in 2021 came back in late 2025 to say they were wrong: once Chinese growth slowed, poor-country growth collapsed because there was no durable capital accumulation underneath it. Mozambique’s per capita growth fell from 5.2% to 0.3%, Zambia from 3.5% to 0.3%.
If you look at the profession as a whole, you see a lot of economists pointing at this gap for decades. They are routinely sidelined in favor of voices whose concern for market discipline ends where Wall Street’s balance sheet begins.
Where I’m going with this
I’m taking Cass’s and Baker’s arguments in a more “developmental state” direction, with pensions being the biggest target. Three claims:
First, the capital that energy, infrastructure, and housing buildouts need already exists: $6 trillion in public pension funds, with liability profiles that match 30-year transmission lines and 40-year nuclear plants almost perfectly.
Second, that capital is trapped in an institutional arrangement that charges enormous fees to match the performance of a simple index fund, while the money deployed through these intermediaries actively harms the communities pensioners live in: buying up their housing, closing their hospitals, bankrupting their local employers, gutting their newspapers.
Third, redirecting this capital doesn’t require new government spending or new financial structures. It requires breaking the equilibrium that one side of the debate defends (Summers) in favor of the other (Baker, or further still, the developmental-state economists in China, Singapore, and Japan). And it requires conditioning any future public rescues on actual structural reform.
We want to start with the data, because the data are damning.
The Center for Retirement Research at Boston College ran a straightforward comparison: how did actual public pension returns stack up against a simple 60/40 portfolio of stock and bond index funds, the kind any individual investor could assemble in a Vanguard account for nearly zero cost? The answer: roughly the same. Public pensions earned about 6.1% annually. The 60/40 benchmark earned about 6.1% annually. The entire apparatus of consultants and alternative managers produced no excess return.
It gets worse. Pensions performed better than the benchmark before the 2008 financial crisis and worse afterward, precisely as they ramped up allocations to hedge funds, private equity, and other “alternatives.” Richard Ennis has documented a negative alpha of approximately 1.2% per year since 2008, virtually all of it associated with alternative investments. Reason Foundation found that 84% of public pension funds failed to beat a passive 60/40 portfolio over the past 20 years. The median plan trails the benchmark by 1-2% annually over 15- and 20-year horizons.
Cass documents the same pattern from the other end of the transaction: private equity is now underperforming a simple S&P 500 index fund across three-month, one-year, three-year, five-year, and ten-year horizons.
The fee structures explain why. For a large pension fund with 30-40% in alternatives, total portfolio costs exceed 1% annually, versus roughly 5 basis points for an indexed strategy. The most expensive slice is private equity, where Ennis estimates total costs of 6%+ annually. On $6 trillion in assets, 1% is $60 billion per year. Over a decade, half a trillion dollars. Over 30 years, more than a trillion.
CalPERS, the nation’s largest public pension, exited hedge funds in 2014, citing “high costs and weak performance.” Their hedge fund portfolio had generated 4.8% annually over a decade, trailing the hedge fund industry’s own benchmark by roughly a fifth. Then, by 2024, CalPERS announced plans to increase private market investments by $30 billion. The cycle repeats because nothing about the incentive structure has changed.
The tens of billions in annual fee waste is actually the smaller problem.
The larger problem is what pension capital is and what it’s being used for. Pension capital is money that will not be needed for 20 to 40 years. By its nature, it is the most patient capital in the American economy, longer in duration than venture capital (5-7 year exit targets), longer than private equity (similar horizon despite the “long-term” branding). Pension funds should be the natural financiers of the long-duration assets this publication covers: transmission lines, nuclear plants, housing bonds, grid infrastructure.
Instead, pension funds hand their capital to intermediaries who invest it on shorter time horizons and charge 6% annually. The assets that need patient capital, grid, nuclear, housing, defense-industrial capacity, can’t find it. And the intermediaries collect fees regardless.
We’ve solved this kind of problem before.
What we are describing is not novel. It is approximately what American public pensions did for four decades before the alternatives revolution.
From roughly 1920 to 1960, legal restrictions confined pension investments to government securities, primarily municipal and state bonds. Michael Glass and Sean Vanatta call this “fiscal mutualism.” At the system’s peak in 1942, public pension portfolios held 73% of their assets in state and local bonds. New York Comptroller Morris Tremaine invested pension funds in municipal bonds explicitly “to assist local governments in procuring funds for improvements at fair interest rates.” The bonds paid market rates. This was not a subsidy. But the closed loop aligned the interests of pensioners, taxpayers, and local governments in a way the current system does not. A retired firefighter’s savings backed the fire station. There were essentially no intermediary fees.
The mechanism generated what Glass and Vanatta describe as a “virtuous cycle.” Public employees contributed wages to the pension fund. The fund bought municipal and school district bonds. The captive demand suppressed borrowing costs, allowing localities to build schools, roads, and sewers at favorable rates. That infrastructure supported community growth, which expanded the tax base and the number of public employees, which increased contributions to the pension fund. Nobody got rich intermediating the transaction, which is why nobody has an incentive to bring it back.
The system started to strain in the 1940s when municipal bond yields fell to around 1.2%, roughly 41% of corporate bond returns, and broke open in the 1960s when states adopted the “prudent man rule” allowing pension funds to invest in corporate equities. That first shift, from municipal bonds to corporate stocks, was defensible. Equities offered higher long-term returns, and pensioners deserved the benefit. The adoption of the prudent man rule by states and its role in transforming public pension investment is documented in detail by Vanatta.
The second shift was not. Starting in the 1980s and accelerating after 2000, pensions piled into hedge funds, private equity, and other alternatives that promised superior returns through sophisticated strategies. By 2021, alternatives constituted nearly 40% of pension risk allocations. Two decades of data now show this shift was a pure loss: same returns, vastly higher fees, and no benefit to the communities whose workers the funds are supposed to serve. The virtuous cycle was replaced by an extractive one. Pension contributions flow to consultants, who recommend alternative managers, who charge 6% in fees and deliver index-matching returns. Meanwhile the communities that employ the pensioners struggle to finance basic infrastructure.
We are not arguing for a return to 1942. We are arguing that some version of the closed loop should replace the current arrangement.
Japan and China both built their own versions of fiscal mutualism at national scale.
Japan ran the Fiscal Investment and Loan Program (FILP) from the postwar period through 2001, channeling postal savings and pension reserves through the Ministry of Finance into infrastructure and public housing. At its peak, the system controlled assets equivalent to 82% of GDP. But it was politically captured: by the 1990s, the Liberal Democratic Party was using FILP funds for the Daini Tomei freeway, an $83 billion project to parallel an existing highway with declining traffic.
Japan reformed the system in 2001, creating the Government Pension Investment Fund (GPIF), now the world’s largest pension fund at $1.8 trillion. GPIF holds just 1.6% in alternatives, compared to 30-40% for American public pensions. The fee gap tells the story. GPIF pays its external managers roughly 2 to 4 basis points of portfolio value; in its record fee year, total payouts were ¥61 billion on a ¥186 trillion portfolio, about 3.3 basis points. American public pensions report average investment management expenses of 39 basis points, but that dramatically understates the real number: when unreported alternative investment fees including carried interest are counted, the full cost is closer to 100 basis points. American pensions pay 25 to 50 times more in fees than GPIF. And the extra spending doesn’t buy anything. GPIF has earned 4.33% annualized since 2001. American public pensions have earned roughly 7-8% annualized over the same period, higher in nominal terms and incredible variation of results from plan to plan, but Boston College’s Center for Retirement Research found that since 2000, pension fund returns have been “virtually identical to a simple 60/40 portfolio” of index funds. The move into expensive alternatives added cost without adding return. The Japanese failure was political capture of capital allocation. The American failure is intermediary capture, which may be even worse. Both are real risks, and any reform must guard against both.
But here is what Japan got right that America is getting wrong. As of December 2024, 88.1% of Japanese government debt is held domestically. The Bank of Japan holds 46.3%, domestic insurance companies 15.6%, domestic banks 14.5%. Foreign investors hold only about 12%. Japanese households save at high rates, deposit into banks and insurance companies, and those institutions buy Japanese government bonds because their liabilities are also yen-denominated. The government uses the proceeds for spending and debt service. Even after the formal FILP reforms in 2001, the structural logic persists through this market-mediated channel. The institutional successor is hiding in plain sight. Japan Post Bank, the semi-privatized descendant of the postal savings system that once fed FILP, still holds ¥190 trillion in deposits (~$1.29 trillion), still maintains ¥40.3 trillion in Japanese government bonds, and still operates across nearly 24,000 branches serving 120 million accounts. It posted record profits for the second consecutive year in FY2025. The government still holds a 22% economic interest. The fiscal mutualism did not disappear with the 2001 reform. It became more market-mediated, for better or worse. Japan Post Bank channels household savings into government debt and, increasingly, into a diversified global portfolio of ¥87.4 trillion in foreign securities. The postal savings pipeline that built postwar Japan is still operating. It just has a stock ticker now.
Japan carries gross debt exceeding 230% of GDP, levels that would trigger capital flight in any country dependent on foreign creditors. It hasn’t. The domestic savings channel has proved so resilient that betting against JGBs became known on Wall Street as the “widowmaker trade.” Every macro fund that shorted Japanese government bonds on the theory that the debt was unsustainable lost money, because the theory ignored the closed loop of domestic savings financing domestic debt.
The widowmaker shorts were wrong for a deeper reason than they realized. The domestic holding base doesn’t merely prevent capital flight. It keeps funding costs so low that the entire Japanese public sector operates (as we are clearly seeing) what economists at the St. Louis Fed and Stanford have called “a sovereign wealth fund with borrowed money.” Domestic savers accept low yields on JGBs because their liabilities are yen-denominated and the alternatives are limited. The government borrows at those low rates and holds the proceeds (directly and through entities like GPIF and the Bank of Japan) in higher-returning assets: foreign reserves, equities, and other risky positions. Research by YiLi Chien of the St. Louis Fed, Wenxin Du of Harvard, and Hanno Lustig of Stanford estimates that this strategy has earned the Japanese government an additional 6 percent of GDP per annum above its funding costs over the past decade.
The scale becomes concrete when you look at the individual channels. The Bank of Japan holds ¥83.2 trillion ($532 billion) in exchange-traded funds, making it the largest single shareholder in the Japanese stock market, owning an estimated 7% of the Tokyo Stock Exchange’s market capitalization. Those holdings carry ¥46 trillion in unrealized gains, a 124% return on cost, and generate ¥1.2 trillion in annual dividends that flow to the national treasury. When the BoJ announced in September 2025 that it would begin selling these holdings, the divestment plan was set at ¥625 billion per year, a pace that would take over a century to complete. While other major central banks sit on underwater bond portfolios, the BoJ’s equity book is ¥44 trillion in the black. Add Japan’s foreign exchange reserves, $1.39 trillion as of January 2026, mostly in dollar-denominated assets earning the rate differential over near-zero yen funding costs. Japan’s central bank is running a profitable equity long book funded by money creation, its finance ministry is running a leveraged carry trade in foreign currencies, and its pension fund is compounding returns at minimal cost. The “sovereign wealth fund with borrowed money” is not a metaphor. It is a description of actual positions totaling several trillion dollars.
This is why the gross debt number is misleading in isolation. When you consolidate the entire public sector balance sheet (the Bank of Japan’s equity holdings, the foreign exchange reserves, GPIF, the public financial institutions) Japan’s net public sector liability had fallen to roughly 65 percent of GDP by the third quarter of 2025. When the St. Louis Fed researchers ran the same calculation for the United States in 2022, America’s net public sector liability was 119 percent of GDP, nearly double Japan’s current figure, despite America’s gross debt being barely half of Japan’s in GDP terms. The country that looks fiscally reckless on the headline number is in better net shape than the country that looks responsible, because Japan kept its savings at home and earned a spread on them while America scattered its capital across fee-extracting intermediaries.
Japan is not the only country running this playbook. Singapore carries gross debt of 173% of GDP, higher than the United States, but has no net debt. Its sovereign wealth funds, GIC and Temasek, together with the Monetary Authority of Singapore manage assets estimated at three to four times GDP. Investment returns from these reserves contribute roughly 20% of government revenue, about 7% of GDP annually. Singapore’s constitution bars the government from borrowing to spend; instead it allows spending up to half of expected long-term real returns on the reserves. As OMFIF put it: “Singapore has employed borrowing not to meet current expenses or to fund a deficit, but to build a strong balance sheet.” The crude debt-to-GDP metric that dominates American fiscal debate would rank Singapore as more indebted than the United States. In reality it is one of the wealthiest sovereigns on earth.
The principle is the same in both cases. The domestic holding base is the load-bearing wall of fiscal stability; the direct product of keeping capital circulating inside the national economy rather than handing it to intermediaries who scatter it across global alternative assets. The St. Louis Fed researchers note that Japan’s ability to sustain this depends on “stable and low funding costs from bondholders,” meaning the domestic savings channel must hold. If domestic institutions stop buying JGBs, the spread collapses and the sovereign wealth fund logic unravels. America cannot copy Japan’s solution or Singapore’s. But the underlying principle is clear: a nation that keeps its savings circulating domestically earns the fiscal room to maneuver that a nation dependent on foreign creditors and fee-extracting intermediaries does not.
China built a similar engine through the China Development Bank and Local Government Financing Vehicles, routing household deposits into infrastructure at massive scale. The CDB is the world’s largest development bank, larger than the World Bank, with RMB 18.2 trillion in total assets (~$2.5 trillion). It is the second-largest bond issuer in China after the Ministry of Finance, and its debt is treated as risk-free under Chinese capital adequacy rules. CDB issues sovereign-status bonds, raises capital from commercial banks at low rates, and lends to strategic projects: Three Gorges Dam, Shanghai Pudong Airport, the high-speed rail network, the Belt and Road Initiative.
The physical output is visible at continental scale. China’s high-speed rail network reached 48,000 kilometers by the end of 2024, roughly two-thirds of all high-speed rail on earth, built in about sixteen years from essentially zero. China built it at roughly one-third the per-kilometer cost of comparable Western projects — California’s unfinished HSR runs around $52 million per kilometer, while China’s 350 km/h lines cost $17-25 million. Only about 6% of the network is profitable in narrow financial terms, but a World Bank study found an 8% annual economic return when accounting for travel times, labor mobility, regional development, and reduced congestion. Fiscal mutualism optimizes for national returns, not financial returns to intermediaries.
LGFVs used land as collateral to borrow for local infrastructure, physically transforming the country’s built environment in a single generation. By the early 2020s, LGFV debt had reached an estimated $8.3 trillion, with more recent estimates placing total liabilities at roughly RMB 78 trillion (~$10.8 trillion), and the property crash exposed it as a potential debt trap.
China’s response in 2024 was to swap LGFV debt into official sovereign bonds. The State Council secured approval for ¥6 trillion in new local government debt limits plus ¥4 trillion from special-purpose bonds over five years, totaling ¥10 trillion in direct debt reduction. The interest rate arbitrage is the point: some weak LGFVs had been paying effective rates as high as 8-16%, while sovereign local government bonds carry rates of roughly 2-3%. The swap is a massive transfer of interest expense from local governments to the sovereign balance sheet — the same logic Japan applied through its domestic holding structure, executed at a different scale and speed.
But the more important development for our purposes is what China is doing next with its state-mobilized capital. The National Social Security Fund has pivoted toward domestic semiconductor and artificial intelligence companies as part of a whole-of-nation effort at technological self-sufficiency. The most direct instrument is the National Integrated Circuit Industry Investment Fund Phase III, known as Big Fund III, established in May 2024 with ¥344 billion ($47.5 billion), the largest state-backed semiconductor fund in history. It is backed by 19 state-owned investors led by the Ministry of Finance and CDB Capital, with six major state-owned banks providing additional capital. Where previous phases focused on fabrication capacity, Big Fund III targets the supply chain chokepoints that US export controls have identified as vulnerabilities: lithography tools, inspection systems, etching platforms, photoresists, and wafer materials. Across all three phases, the Big Fund has mobilized roughly $95 billion, nearly double the US CHIPS Act’s $53 billion in incentives. Domestic savings flow through state banks into the Ministry of Finance, through CDB Capital, and into the industrial base that will determine whether China can manufacture semiconductors without foreign permission.
The comparison is hard to ignore. China directs state-mobilized capital toward strategic technology competition and builds physical infrastructure at a third the cost, financed through a banking system that treats development lending as a sovereign function. Japan runs a disciplined, low-cost portfolio and keeps its domestic savings circulating where they underwrite fiscal stability and a 6 percent of GDP annual spread. Singapore borrows at 173% of GDP and funds a fifth of government revenue from the proceeds. America pays tens of billions per year to financial intermediaries who add no value, while the infrastructure and defense-industrial capacity that pension capital should be financing goes underfunded. We don’t need to emulate China’s state-directed model. But we should stop being the only major economy whose largest pool of patient capital is optimized for fee extraction rather than national capacity.
So where would this capital actually go?
The housing case is the most proven, because it’s already working at small scale.
New York City’s pension systems allocate 2% (but effectively less) of assets to economically targeted investments, deploying $852 million through the AFL-CIO Housing Investment Trust since 2002 and financing over 37,000 affordable apartments using union labor. The GAO evaluated these programs and found they achieved competitive returns.
Texas offers another model at the infrastructure level. Municipal Utility Districts are special districts that issue tax-exempt bonds to finance water, sewer, drainage, and road infrastructure for new developments. MUD bonds are self-liquidating (homeowners pay them off through property taxes over 20-30 years), asset-backed, transparent, and regulated by the Texas Commission on Environmental Quality. They have financed major projects like The Woodlands, Clear Lake City, and First Colony. Pension funds holding MUD bonds instead of hedge fund LP interests would earn comparable returns while financing actual housing and infrastructure.
The question is why New York’s ETI allocation is 2% instead of 15%. The answer is institutional inertia and the same consultant-driven incentive structure that pushes everything toward complex, high-fee alternatives. The mechanism works. It is just not being used at scale.
mattparlmer argued that America can beat China on energy, but identified the “location problem” as a critical bottleneck: cheap power exists in places like Texas and the Pacific Northwest but can’t reach population centers like San Francisco or New York. His conclusion: “New transmission capacity must be heavily subsidized.”
He’s right, and we need to go further with providing cheaper capital. Transmission lines are 30-50 year assets with regulated, predictable cash flows. They don’t need charity, they need patient buyers for the debt that finances them. The natural buyers are pension funds, whose 30+ year liability profiles match transmission asset durations almost perfectly. The reason pension funds aren’t buying is that their capital is tied up in alternative strategies with 5-7 year horizons, paying 6% in annual fees to roughly match a bond index.
Pension funds currently hold around $2 trillion in the bond side of their portfolios. Most of that is in corporate bonds, complex fixed-income strategies, and private credit funds charging 2-and-20. If even a quarter of that allocation shifted to municipal and infrastructure bonds, including bonds financing grid transmission, that would create roughly $500 billion in captive demand for infrastructure debt. That would lower the cost of financing transmission buildout. Not through subsidies. Through a large, natural buyer entering the market for the assets its liability structure already demands.
Parlmer estimates that solving the location problem might be part of a greater-than-$1 trillion project. The pension system’s bond allocation alone could absorb a significant fraction of that financing need, if it weren’t locked up in hedge fund LP interests.
China has roughly 30 reactors under construction, financed in significant part through state-mobilized patient capital channeled via the China Development Bank. American pension funds hold $6 trillion in equivalent long-duration capital and are spending it on hedge fund fees.
Thomas Hochman has written about the LPO’s role in nuclear financing and the double-dipping restrictions that impede it. The LPO is a critical tool, providing federal debt for projects that private lenders consider too large or technically complex. But nuclear projects also need long-term equity commitments and project finance bonds that match their 40-60 year operational life. A nuclear plant with a power purchase agreement is essentially an infrastructure bond: predictable cash flows over a multi-decade horizon, backed by a regulated revenue stream. Pension funds should be natural participants in that capital stack.
Parlmer has floated the idea of state-backed reactor fleets, either DoD or DoE operated, because he can’t see another path to rapid deployment. Pension-backed project finance isn’t state ownership, but it could serve a similar de-risking function. A pension fund investing in a nuclear project finance bond backed by a DoD power purchase agreement would face lower risk than most of what is currently in pension alternative allocations, and would fund something that matters.
Parlmer’s analysis of Chinese industrial automation is alarming: PRC added more industrial robots in 2023 than the rest of the world combined, quintupled its per-worker robot count since 2017, and is pulling ahead of the United States in industrial sophistication. His prescription: dollars flowing into industrial automation at more than ten times current levels, with subsidies directed to startups rather than legacy firms.
Pension funds already allocate significant capital to venture through their alternatives sleeve, but that money flows overwhelmingly to Sequoia and Andreessen, not to defense tech or industrial automation startups in Ohio or Texas. The geographic and sectoral concentration of pension-funded VC mirrors the broader venture ecosystem’s biases: software over hardware, Bay Area over everywhere else, consumer over defense.
Redirecting even a fraction of pension VC allocation toward regional defense and industrial venture vehicles wouldn’t require new government spending. It would redirect existing capital flows. The Ohio Innovation Fund and similar state-level vehicles demonstrate that the institutional architecture is feasible. The returns are uncertain (this is venture, after all) but the counterfactual is not a guaranteed 12% from private equity. It’s 6% from a fund that charges 6% in fees to roughly match the S&P 500.
We will be candid: the venture case is weaker than the infrastructure case. Pension funds are not natural venture investors, and pushing them toward VC risks the “picking winners” failure mode that destroyed Japan’s Fiscal Investment and Loan Program. But our argument is not that pensions should become venture funds. It is that the capital they currently waste on underperforming alternatives could be partially redirected toward asset classes that at least fund something strategic, and that the bar for improvement over the status quo is extraordinarily low.
Pension capital allocated to private equity and hedge funds does not just fail to generate excess returns. It actively harms the communities pensioners live in: their housing, their hospitals, their local employers, their newspapers. The capital that once built those communities through municipal bond investment now funds their extraction. Public pension funds provide roughly one-third of all private equity investment capital, over $620 billion in 2022, up from 3.5% of pension assets in 2001.
Start with Chelmsford, Massachusetts. Public school custodians earning $25 an hour had their jobs privatized when Aramark, then owned by private equity and backed by 37 state and local retirement funds, underbid their union for the school cleaning contract. Aramark offered the custodians their jobs back at $8.75 an hour, a 56% pay cut. Their own pension savings funded the company that destroyed their livelihoods.
Chelmsford is not an outlier. Over the past decade, approximately 597,000 workers at PE- and hedge fund-owned retailers have lost their jobs directly. When indirect supply-chain effects are included (the Economic Policy Institute estimates 122 additional jobs lost per 100 direct retail losses), the total approaches 1.3 million. Toys “R” Us, 99 Cents Only Stores, Joann, Big Lots, Red Lobster: PE-owned firms go bankrupt at roughly ten times the rate of comparable non-PE firms. Cass catalogues the pattern across veterinary practices, funeral parlors, campgrounds, youth sports, and volunteer fire departments, all consolidated and squeezed and repackaged as “value creation.”
In local media, hedge fund Alden Global Capital and similar firms now control more than half of America’s daily newspaper circulation. Over 2,000 local newspapers have closed nationwide. When local papers disappear, research shows communities experience increased political polarization, decreased civic engagement, and higher borrowing costs for local governments, because the loss of journalistic oversight reduces transparency. The function that fiscal mutualism once served, lowering government borrowing costs through captive pension demand, is being actively undermined by what pension capital currently funds.
Cass argues that financialization should be treated as “a grift, a rarefied form of bookmaking.” But the Chelmsford custodians don’t need a cultural reckoning. They need their pension fund to stop financing the company that cut their wages by 56 percent.
Steward Health Care shows how the extraction works at its most lethal. In 2010, Cerberus Capital Management bought Caritas Christi, a Catholic hospital system in eastern Massachusetts, for $895 million, putting up only $246 million in equity. Cerberus subsequently sold Steward’s hospital real estate to Medical Properties Trust for $1.25 billion in a sale-leaseback that saddled the hospitals with unsustainable rent obligations. All told, Cerberus and Steward CEO Ralph de la Torre extracted approximately $1.3 billion from the system. De la Torre bought a $40 million yacht and two private jets. Steward filed for bankruptcy in 2024 with $9 billion in liabilities. Five hospitals closed. Roughly 2,400 workers lost their jobs. Nashoba Valley Medical Center, which served 115,000 residents in rural Massachusetts and received 16,000 emergency room visits annually, shut its doors. First responders now travel roughly 15 miles to the next hospital.
Steward is not an outlier either. A 2024 Review of Financial Studies paper found that private equity acquisition of nursing homes was associated with an ~10% increase in deaths, implying approximately 22,500 additional deaths over the twelve-year sample period. A Harvard Medical School study the same year found PE ownership associated with a 25% increase in hospital-acquired conditions. A 2021 JAMA study found nursing home residents under PE ownership 11% more likely to visit emergency departments and 9% more likely to be hospitalized for preventable conditions. No study to date has found PE ownership improves patient outcomes or lowers costs.
Private equity investment in healthcare has grown from $5 billion in 2000 to an estimated $104 billion in 2024. PE firms now own approximately 488 U.S. hospitals, more than one in five for-profit hospitals, and roughly 40% of emergency departments are staffed or managed by PE-owned companies.
The public money doesn’t just flow to private equity and hedge funds through pension fees. It flows through bailouts too, and in both directions.
Start with hedge funds. In 1998, the Federal Reserve brokered a $3.625 billion rescue of Long-Term Capital Management, coordinating 14 banks to recapitalize a single hedge fund whose collapse threatened the broader financial system. That set a template. In March 2020, when COVID triggered a liquidity crunch, hedge funds running highly leveraged Treasury basis trades started dumping Treasuries en masse. The Fed responded by purchasing roughly $1.6 trillion in Treasury securities over several weeks. The basis trade was effectively backstopped with public money. That trade has since doubled in size, to roughly $1 trillion, concentrated among fewer than ten hedge funds. And in March 2025, a Brookings Institution paper co-authored by former Fed governor Jeremy Stein proposed formalizing this arrangement: a standing “basis purchase facility” that would allow the Fed to unwind hedge fund positions during the next crisis. The financial establishment is now pre-planPublic money doesn’t just flow to private equity and hedge funds through pension fees. It flows through bailouts too, and in both directions.
Start with hedge funds. In 1998, the Federal Reserve brokered a $3.625 billion rescue of Long-Term Capital Management, coordinating 14 banks to recapitalize a single hedge fund whose collapse threatened the broader financial system. That set a template. In March 2020, when COVID triggered a liquidity crunch, hedge funds running highly leveraged Treasury basis trades started dumping Treasuries en masse. The Fed responded by purchasing roughly $1.6 trillion in Treasury securities over several weeks. The basis trade was backstopped with public money. That trade has since doubled in size to roughly $1 trillion, concentrated among fewer than ten hedge funds. And in March 2025, a Brookings Institution paper co-authored by former Fed governor Jeremy Stein proposed formalizing this arrangement: a standing “basis purchase facility” that would let the Fed unwind hedge fund positions during the next crisis. The financial establishment is now pre-planning the next hedge fund bailout.
Now the pension side. Years of fee extraction and underperformance have left pensions deeply underfunded, $1.3 trillion in the hole by official accounting, or closer to $5.1 trillion using market-rate discount rates, according to Stanford’s Joshua Rauh. That underfunding eventually triggers its own bailouts. In 2021, the American Rescue Plan included roughly $97 billion in “special financial assistance” for the most severely underfunded union pension plans. The single largest allocation, $36 billion, went to the Teamsters’ Central States Pension Fund. That assistance has improved headline numbers: Milliman’s year-end 2024 study reports aggregate multiemployer funding at 97%. But strip out the SFA grants and the aggregate drops to 89%. Eighty-five plans remain below 60% funded. And the SFA program made no changes to multiemployer plan funding rules, so the structural dynamics that created the underfunding are entirely intact.
The obvious objection to redirecting pension capital is political capture, Japan’s FILP and its bridges to nowhere. That risk is real. But Japan’s pension system, whatever FILP’s failures, is not trillions in the hole. The counterfactual is not “safe status quo vs. risky reform.” The current system already fails, already gets bailed out, and taxpayers are already on the hook.
On top what the pension funds already pay out in fees, if the federal government is going to spend $97 billion rescuing pension plans, and the Fed is going to deploy trillions backstopping the Treasury market, the obvious question is: why isn’t any of that capital building something?
A $97 billion pension rescue conditioned on not just reforms but also investing 25% of assets in domestic infrastructure bonds would have simultaneously stabilized the pensions, financed grid buildout, and created a template for redirecting patient capital toward national priorities. Instead, the money went in with no structural reform. No investment mandates, no fee caps, no requirement to stop feeding the same intermediary complex that created the underfunding. The plans that received bailout funds are free to continue allocating to the same managers whose fees helped drain them. And they will.
The institutional dynamics are a textbook captured-bureaucracy problem. Career risk provides cover through complexity: run a simple indexed portfolio and underperform in a given year, and you look unsophisticated. Run a complex alternatives-heavy portfolio and underperform, and you look like you tried your best. No one gets fired for hiring Goldman Sachs. Consultants don't get paid to recommend index funds; they get paid to recommend strategies requiring ongoing advice and manager searches. Benchmarks are gamed: pension funds construct custom benchmarks biased by 1-2% annually relative to fair passive alternatives, and private equity returns are reported with lags that smooth volatility and create the illusion of diversification. Trustees worry that directing capital toward municipal or infrastructure bonds rather than "optimizing" globally could be challenged as a breach of fiduciary duty, even though the "safe" legal path leads into strategies that have demonstrably underperformed. The cycle is self-reinforcing. CalPERS tried to break it in 2014 when they exited hedge funds. Ten years later, they were back in private markets.
The honest version of this section would list transparency mandates, fee caps, infrastructure bond requirements, and indexing mandates. We drafted it. Then we deleted it, because every item on the list requires action by the same actors who benefit from the current arrangement.
State legislators who would mandate fee transparency receive campaign contributions from the asset management industry. Pension board trustees who would cap alternatives face career risk if they simplify. The consultants who would need to recommend index funds over complex strategies are the same consultants whose business model depends on recommending complex strategies. Federal legislators who would attach conditions to bailouts just passed $97 billion with no conditions attached, because the political incentive was to deliver the money and move on.
We have spent this entire piece documenting a self-reinforcing capture loop. It would be dishonest to end it by proposing that the captured institutions reform themselves. CalPERS knew the answer in 2014. They exited hedge funds, published the rationale, took a public stand. A decade later they were back in private markets for $30 billion more, because nothing about the incentive structure had changed. If the nation’s largest and most sophisticated pension fund, with a board that understood the problem and acted on it, could not sustain the exit, a policy white paper is not going to do it.
We are describing a trap, not proposing an escape.
What will happen is another bailout. The math guarantees it. The SFA grants improved headline numbers, but strip them out and aggregate multiemployer funding drops to 89%. Eighty-five plans remain below 60% funded. The program changed no funding rules. The fee extraction continues. The underperformance compounds. At some point, probably within the next decade, a recession or market downturn will force the next round of plans into insolvency and Congress will face the same choice it faced in 2021.
That is the only moment when the political dynamics actually shift. When public money is on the table and legislators need a story to tell about why this time is different, there is a narrow window in which conditions can be attached. Not because anyone in the system wants reform, but because the politics of unconditional bailouts get worse each time. The 2021 rescue passed in a COVID relief omnibus where no one was paying attention. The next one will face more scrutiny, more anger, and a public that has watched pension-funded private equity close their hospitals and cut their neighbors’ wages.
The question is not “how do we reform pension allocation now.” It is: “when the next bailout arrives, what conditions are ready to attach?” That means having the legislative language drafted. It means having the qualifying criteria for infrastructure bonds defined (investment-grade, revenue-backed, independently certified) so that the debate is over a concrete mechanism rather than an abstract principle. It means having the fee cap number chosen and the benchmark methodology specified. It means having the sunset clause and the performance trigger designed, so that the infrastructure mandate self-destructs if it fails and cannot become a permanent slush fund.
None of that requires passing legislation today. It requires preparation for a political window that is coming whether anyone wants it or not.
The trap will not reform itself. But it will need public money again. The work between now and then is to make sure that the next time Congress writes a check to rescue a pension system bled dry by intermediary fees, the check comes with conditions that route the capital toward qualifying infrastructure debt rather than back into the same extraction loop that created the crisis. That is the difference between rescue and subsidy. It is also, realistically, the only reform that has a chance of happening.