Two years ago, I and John Mantus, and Gaobo Pang (collectively, “WMP”) created a long-range macroeconomic growth simulation model to project the fiscal condition of the US government under current broad policies and economic trends. Although similar in many of its assumptions and methodologies to those used by the Social Security and Medicare Trustees in their annual Reports, and especially to the Congressional Budget Office (“CBO”) in its periodic 10- and 30-year projections, a unique aspect of the WMP model is its explicit inclusion of the production of health care services as a separate sector in a growth model with capital and labor. This addition was done in recognition of the large and growing share that health care plays in the US economy (now approaching a fifth), the significant federal budget burden it represents (covering half of health spending), and especially in light of the inefficiency of the sector: It uses relatively little capital and a lot of labor with low productivity growth, leading to increasing relative prices of health care and lower consumer welfare. In this blog post, I report on the results of the updated model, comparing it to last year’s results and to CBO’s most recent projections. In short, the fiscal outlook is slightly improved because real interest rates are now projected to rise less, yet compared to CBO’s projections, which are based largely on current law and not policy and incorporate some unreasonable assumptions, the outlook is more dire.
The main input changes for the model this year are a lower projected population and labor force, slightly higher initial capital stock, slightly lower initial real interest rate, near-complete elimination of taxation on Social Security benefits, and other small technical fixes. Recent major changes in tax and health care law are reflected only to the extent they appear empirically in broad aggregates and trends. This makes sense because these changes are often temporary (tariffs ruled illegal by the Supreme Court), representative of current policy (2017 tax cuts made permanent last year), offset by behavior responses (tax avoidance or increased health care utilization), countervailing (Medicare enhancements versus Medicaid cuts), or unsustainable (price controls in the health care sector eventually undone by legislation). Unlike CBO, which assumes that defense and non-defense discretionary spending will decline as a share of GDP over time, the WMP model uses more realistic assumptions: Defense spending will remain constant as a share of GDP and non-defense discretionary will scale with the population. Also, CBO and conventional government accounting overstate the deficit by ignoring the implicit inflation tax that erodes the real value of government debt held by the public. The smaller projected labor force now is the key because it reduces demand for capital outside of the health care sector, lowering the real interest rate. Model results for WMP last year and this, compared to CBO’s most recent projections, over the next 10 and 30 years, are shown below.
According to the WMP model, the deficit is projected to increase from about four percent of GDP currently (six percent in conventional terms, ignoring the implicit inflation tax) to 7.5 percent in 10 years. The ratio of federal debt held by the public will rise from about 100 percent currently to 138 percent in 10 years. Beyond that, current policy is clearly unsustainable, with deficits and debt at nearly 14 and 270 percent of GDP, respectively, in 30 years. Even without igniting a crisis, such high levels of debt combined with increased spending on inefficiently produced health care (from 18 percent of GDP currently to almost 26 percent) substantially reduce consumer welfare growth, eventually turning it negative.
The deficit and debt projection in WMP is uniformly higher than CBO’s, controlling for the inflation tax effect, for several reasons. First, real interest rates in WMP are higher, in recognition of current relatively high rates, the competition for funds that increasing deficits represent, and the inefficiency of the health care sector drawing away labor from the productive “all other” sector. Second, the projection for federal benefit spending, which includes Social Security, Medicare, Medicaid, and other federal health programs, is higher in WMP because CBO estimates the effect of current law on federal health spending, which constrains payments to hospitals, physicians, and drug manufacturers unsustainably in the current inefficient health sector environment. CBO also does not seem to recognize the demographic uptick in Medicaid from long-term care, and it assumes that the increase in relative health care prices will subside over time. Third, as mentioned, CBO projects unrealistically low spending on defense and non-defense discretionary items and projects higher tax revenues from bracket creep, which history shows is always effectively undone by legislation.
Facts Only
A team including John Mantus and Gaobo Pang created a macroeconomic growth simulation model two years ago.
The model projects the U.S. government's fiscal condition under current policies and economic trends.
The model uniquely includes healthcare as a separate sector, recognizing its large and growing share in the economy.
The updated model shows a slightly improved fiscal outlook due to lower projected real interest rates.
The model projects the deficit to increase from about 4% of GDP currently to 7.5% in 10 years.
Federal debt held by the public is projected to rise from about 100% of GDP currently to 138% in 10 years.
In 30 years, deficits and debt are projected to reach nearly 14% and 270% of GDP, respectively.
Healthcare spending is projected to rise from 18% of GDP currently to almost 26% in 30 years.
The model's projections are higher than the CBO's due to higher real interest rates and higher federal benefit spending.
The CBO assumes unsustainable constraints on healthcare payments and underestimates Medicaid costs for long-term care.
The WMP model projects stable defense spending and population-scaled non-defense discretionary spending.
The CBO projects declining defense and non-defense discretionary spending as a share of GDP over time.
Executive Summary
Two years ago, a team including John Mantus and Gaobo Pang developed a macroeconomic growth simulation model to project the U.S. government's fiscal condition under current policies. Unlike traditional models, theirs explicitly includes healthcare as a separate sector, reflecting its outsized economic and budgetary impact. The updated model shows a slightly improved fiscal outlook due to lower projected interest rates, but it remains more pessimistic than the Congressional Budget Office (CBO) projections, which rely on current law and assumptions the authors consider unrealistic.
Key differences include the WMP model's higher real interest rates, driven by competition for funds and healthcare inefficiencies, and its projection of higher federal benefit spending, particularly in healthcare. The CBO assumes unsustainable constraints on healthcare payments and underestimates Medicaid costs for long-term care. The WMP model also projects stable defense spending and population-scaled non-defense discretionary spending, contrasting with the CBO's declining projections. Over 30 years, the model forecasts unsustainable deficits and debt, with healthcare spending rising to nearly 26% of GDP, significantly reducing consumer welfare.
Full Take
The strongest version of this narrative highlights the structural inefficiencies in the U.S. healthcare sector and the unsustainability of current fiscal policies. The authors deserve credit for explicitly modeling healthcare's role in economic growth, a critical oversight in many projections. Their critique of the CBO's assumptions—particularly around healthcare cost controls and discretionary spending—is well-founded, as historical trends often undermine legislative constraints.
However, the analysis may overstate the inevitability of its projections. The assumption that healthcare inefficiencies will persist unchecked ignores potential policy innovations or technological disruptions that could improve productivity. Additionally, the model's reliance on current trends risks underestimating adaptive behaviors, such as shifts in labor force participation or healthcare utilization patterns. The framing of the CBO's projections as "unrealistic" could be seen as a strawman, given that the CBO's mandate is to project based on current law, not speculative policy changes.
Root cause: The narrative reflects a broader debate about the sustainability of entitlement programs and the role of government in healthcare. It assumes that inefficiencies in healthcare are intractable without addressing underlying incentives or systemic reforms. This echoes historical patterns where rising costs in critical sectors lead to fiscal crises, but it also risks fatalism by downplaying human agency in policy design.
Implications: If accurate, the projections suggest a future where high debt and healthcare costs erode consumer welfare, limiting economic mobility and public investment. The primary beneficiaries of this narrative are policymakers and economists advocating for structural reforms, while the costs fall on future taxpayers and beneficiaries of federal programs. Second-order consequences could include reduced trust in government institutions if fiscal constraints force abrupt policy shifts.
Bridge questions: What policy levers could mitigate healthcare inefficiencies without relying on unsustainable cost controls? How might technological advancements in healthcare delivery alter these projections? What historical examples of fiscal consolidation could inform a more optimistic path forward?
Counterstrike scan: A coordinated influence campaign pushing this narrative might emphasize fear of fiscal collapse to justify austerity measures or healthcare privatization. However, the content here focuses on analytical rigor rather than emotional manipulation, aligning more with genuine policy debate than a manipulative playbook. The authors' transparency about assumptions and limitations further distinguishes this from a bad-faith argument.
Patterns detected: none