United States Of America map on white background. Tiered 3d render of empty USA territory. Country poster for travel materials, print, banner and web. Geographic area levels visualization.
The U.S. R&D landscape as a whole looks fairly resilient. That’s the finding from an exploration of 10,000 scenarios in a simulations that aims to capture the interplay between federal and private-sector funding under economic strain.
In any event, as 2025 unfolds amid economic uncertainty and budget cuts and a string of R&D related job cuts, the resilience of the U.S. R&D ecosystem faces a stress test. On the one hand, tens of thousands of STEM workers across the country have recently lost their jobs while government science funding faces budgetary pressures. On the other, China is catching up rapidly—when adjusted for purchasing power parity, it may already be near parity with the U.S. in total spending. Either way, China is on track to surge ahead by that metric, as Nature noted.
It doesn’t help that the U.S. faces recent economic challenges, with high uncertainty over a potential recession. Prior downturns provide some clues about what might happen if the U.S. economy continued to deteriorate in 2025. During previous downturns—the dot-com crash, the 2008 financial crisis, and COVID-19—private R&D typically dried up as companies cut or delayed innovative projects. Federal R&D funding, however, historically acted countercyclically, cushioning private cuts through stimulus efforts and stabilizing overall research (See: Recovery Act Helped Maintain U.S. R&D Spending During Economic Crisis | SSTI).
But with deep cuts coming to federal science initiatives, along with substantial layoffs likely in a range of agencies, including the National Science Foundation as well as grant recipients, it remains to be seen whether the broader science landscape could see belt-tightening in the coming years. To explore the possibilities, let’s look at a Monte Carlo simulation, a computational technique that uses repeated random sampling to model uncertain phenomena. The TL;DR? The baseline scenario sees a roughly flatlining of R&D spend from 2025 to 2026, but picking up after that.
10,000 simulations related to the future of U.S. R&D
Our Monte Carlo simulation model was specifically designed to capture the complex interplay between federal and private sector R&D funding during periods of economic uncertainty. The model incorporates 13 parameters including federal funding cuts, mixed sentiment for the semiconductor industry (uncertain CHIPS future vs. increased domestic R&D spending from Big Tech and TSMC’s $165 billion spending plans in Arizona, private sector growth rates, sector-specific changes, and macroeconomic factors. The following model projected a baseline case with a slight dip in R&D spending in 2025 that recovered in the subsequent years.
Monte Carlo models can be useful in exploring potential scenarios. They are sensitive to assumptions, however, and the current R&D landscape is difficult to model.
Second, our model gives substantial weight to tech sector R&D growth (normally distributed with mean 0.06 and standard deviation 0.025) based on strong AI investments that appear resilient even amid economic uncertainty. The widened variance captures the polarity between some companies cutting staff (Meta, Microsoft, Alphabet, etc.) while generally accelerating core AI research. Third, our enhanced correlation logic recognizes that China competition serves as a likely countercyclical force. That is, increased competition from China is likely to spur both a federal and private sector investments increase in response. That creates a partial economic hedge that explains the model’s relatively consistent growth projections through 2027.
In the model, the parameters were weighted based on historical patterns. That is, it assigned higher sensitivity to private sector growth (normally distributed with mean 0.045 and standard deviation 0.015) since it drives roughly 75% of total R&D spending, as NSF has noted.
The model’s correlation logic acknowledges that parameters don’t vary independently. For instance, when interest rates rise significantly, private sector growth is dampened proportionally, and when China competition intensifies, both public and private investment are likely to increase. This approach tends to yield more realistic forecasts than treating each factor separately.
Looking to prior recessions
Current recessionary fears are a wildcard, but prior recessions hold clues about how an economic downturn could impact U.S. R&D spend:
The dot-com bubble impact: The 2001 recession briefly slowed R&D growth in the private tech sector; U.S. business R&D stagnated. Federal R&D (notably defense and health) kept rising modestly, which offset deeper declines, resulting in only a slight dip in total R&D intensity from 2.54% in 2002 to 2.48% by 2004 (See: U.S. R&D Increased by $51 Billion in 2020 to $717 Billion | NSF).
2008–2009 global financial crisis: R&D was initially threatened, but government stimulus cushioned the blow. Total U.S. R&D dipped from $403 billion in 2008 to $400.5 billion in 2009, a 0.6% decline—milder than the 2.5% GDP contraction that year. Business R&D fell 2.9%, while federal R&D rose 3.3%. The American Recovery and Reinvestment Act (ARRA) added billions in federal R&D, causing an 18% surge in federal research obligations in 2009 (ARRA details). This stimulus kept overall research activity from collapsing, and R&D intensity actually rose to ~2.78% of GDP in 2009. China, meanwhile, continued double-digit R&D growth through the crisis, surpassing Japan as #2 in global R&D spending by 2011 (See: China’s spending on R&D rises to new high).
COVID-19 recession (2020): Governments bolstered R&D funding for health and pandemic response even as economies contracted. The U.S. federal R&D obligations spiked 18% in 2020. That was its largest jump since 1963. Vaccine development and medical research helped drive the trend, as NSF has noted. Total U.S. R&D reached $717 billion in 2020 (up $51 billion from 2019), pushing R&D intensity to about 3.4% of GDP. China also bumped up R&D, growing 10.2% year-over-year in 2020 to reach ¥2.4 trillion (~$378 billion), elevating its R&D-to-GDP ratio to 2.4%.
Historically, private-sector R&D is procyclical. During 2008–09, firms with high leverage slashed R&D. Basic and applied research in particular took hits, as NBER has reported. Similar behavior occurred in the early 2000s and two decades later with the initial pandemic shock.
In contrast, government R&D is steadier and can be countercyclical — serving as a sort of teeter totter, historically. As the economy contracts, government R&D has historically expanded. That is, public research budgets, often set in multi-year cycles, are sometimes ramped up for economic stimulus (e.g., ARRA). This offsets part of the private pullback. That occurred in both the Great Recession and the COVID-19 crisis. Yet governments are not immune to cuts: after the 2009 stimulus, sequestration in 2013 curtailed non-defense R&D by about 5%. Still, since business covers approximately 70–75% of U.S. R&D, a private slowdown weighs heavily on total spending.