Public investment draws $2 of private capital for every $1 spent
Study of 34 OECD economies challenges conventional wisdom on public spending effects
In “Invest One, Get Two Extra: Public Investment Crowds in Private Investment,” economists Olegs Matvejevs and Olegs Tkacevs of Latvia’s central bank report a finding that should reshape how policymakers think about government spending. Examining 34 industrialized OECD economies from 1995 to 2019, they conclude that public investment crowds in private investment rather than displacing it. For every dollar governments invest in public capital, approximately two additional dollars of private investment eventually follow.
Conventional Wisdom
For decades, policymakers have assumed that when governments spend more, they compete with private actors for limited savings and labor. Interest rates rise. Private consumption and investment fall. This “crowding out” story has anchored fiscal conservatism and provided a persistent argument against ambitious public spending programs.
The big idea of “crowding out” is that there is only so much savings in an economy. When governments borrow to fund investment, they absorb savings that would otherwise flow to private projects. Increased demand for loanable funds pushes up borrowing costs. Marginal private investments that would have been profitable at lower interest rates become uneconomical. The government builds a bridge; a factory doesn’t get built.
This story isn’t wrong about (very) short-term dynamics. When governments ramp up investment spending, they do initially compete for resources. Interest rates can tick upward. Some private projects get delayed or cancelled. But what happens over the following years matters far more.
What The Research Shows
Matvejevs and Tkacevs employed a “stock-flow” approach that distinguishes their work from studies finding crowding-out effects. Rather than examining only year-to-year changes in investment, they modeled the long-run equilibrium relationship between public and private capital stocks.
Their central finding: public and private capital move together over time. When public capital rises by 1 percent, private capital eventually rises by roughly 0.8 percent. The two stay in rough proportion because public capital makes private capital more productive, which draws in more private investment.
When the ratio gets out of balance, private investment adjusts. If public investment falls short, private investment follows it down. If public investment surges ahead, private investment rises to catch up. A 1 percent gap between the two produces a 1.9 percent correction in private investment the following year.
Consider a logistics company evaluating a new distribution center. The expected return is substantially higher when the center sits near a well-maintained interstate than when it depends on potholed secondary roads. A tech startup benefits from a workforce educated in public universities. A manufacturer’s equipment becomes more valuable when reliable electrical grid infrastructure eliminates costly outages.
Private investment doesn’t simply spike and retreat after a public investment shock. It remains elevated for years as the adjustment process unfolds.
Key Data
The multiplier: Each dollar of public investment generates approximately two dollars of additional private investment over five to seven years.
The cointegration coefficient: A 1 percent increase in public capital stock is associated with a 0.79 percent increase in private capital stock in the long run.
The country range: Multipliers vary from approximately 1.3 in Japan to 4.7 in Latvia.
The decline in public investment: In advanced economies, public investment fell from an average of 2.4 percent of GDP in the 1990s to less than 2 percent after 2010.
Peak timing: The crowding-in effect reaches its maximum around year two after a public investment shock.
The Objections, and Why The Finding Holds
Studies focusing on short-term dynamics, using fixed-effects regressions that don’t account for long-run equilibrium relationships, tend to find crowding out. Studies that model the cointegration between capital stocks find crowding in.
A second objection: perhaps governments invest more when they anticipate private investment will rise anyway, creating a spurious correlation. The researchers argue this reverse causality concern is less severe for public investment than for taxes or transfers. Tax receipts mechanically respond to economic conditions; unemployment benefits automatically rise during downturns. No comparable automatic link exists between private investment plans and public capital spending, which reflects budgetary decisions made through political processes with long lead times.
To address remaining concerns, the authors used public investment forecast errors from OECD Economic Outlook projections. The private sector cannot anticipate public investment more precisely than professional forecasters, so deviations between actual and forecasted investment represent unanticipated shocks. This approach produces positive and significant effects on private investment four to five years out, with modest short-term crowding out in the immediate aftermath.
The findings hold across multiple robustness tests: splitting the sample into pre-2008 and post-2008 periods, restricting analysis to euro area countries alone, substituting IMF data for OECD data, dropping individual control variables, adding lagged controls, including time fixed effects to absorb global shocks.
Several limitations remain. The paper doesn’t address how public investment is financed. Debt-financed investment might eventually raise interest rates enough to offset productivity gains; tax-financed investment might reduce private consumption through other channels.
The research cannot speak to investment quality either. Abiad found crowding-out effects when public investment is inefficient. The multiplier shouldn’t be applied to contexts where waste and corruption dominate.
The sample ends in 2019, before COVID reshaped fiscal policy globally. Whether these relationships hold with higher sovereign debt levels and changed interest rate environments remains untested.
Atukeren, using 25 developing countries, showed that crowding-out likelihood rises with government sector size and capital controls but declines with trade openness. The advanced OECD economies studied here may represent favorable conditions.
Where It Works Best
Not all public investment generates equal crowding-in effects. The study disaggregated spending using the Classification of Functions of Government.
Economic affairs (roads, bridges, transport infrastructure) generates the strongest private sector response. A trucking company’s fleet becomes more productive when highways are maintained. A port expansion makes adjacent warehousing more valuable. Broadband buildouts enable businesses that couldn’t previously exist. Economic affairs account for roughly 35 percent of total public investment across OECD countries.
Education shows similarly strong effects. Public investment in schools and universities produces a more skilled workforce, raising returns to private capital that employs those workers. Education represents about 13 percent of public investment.
Recreation, culture, and religion follows, likely reflecting tourism-related private investment: hotels, restaurants, amenities that cluster around museums, concert halls, and national parks.
Housing and community amenities and social protection show weaker or uncertain relationships. The cointegration evidence for these categories is mixed. Defense spending generates modest effects, perhaps because military capital doesn’t directly raise private factory or office productivity.
The Bottom Line
Many economists (in addition to politicians and fans of industrial policy) have proposed “golden rules” exempting capital spending from deficit targets, reasoning that investments generate future returns while consumption does not.
The multiplier isn’t automatic. It depends on investment type (economic affairs and education outperform housing), execution quality, and whether economies have room to absorb additional spending without bidding up prices. But for governments that invest wisely: one in, two back.