Government investment and fiscal stimulus

Government investment and fiscal stimulus

Fiscal stimulus packages typically feature large investment in infrastructure. The column argues that the fiscal multiplier connected with government investment through the Great Recession was near zero. Meanwhile, the federal government consumption multiplier was around 0.8. Estimates of the multiplier for total government purchases usually do not distinguish both of these effects, which might affect their validity.


Through the Great Recession, governments enacted fiscal stimulus packages to combat the decline in economic activity. Significant shelling out for long-lived investment goods was common to these policies. In america, for example, the American Recovery and Reinvestment Act of 2009 contained provisions to improve funding to spend a lot more than $70 billion on infrastructure and transportation. 1

Far away infrastructure investment was also a substantial fraction of stimulus spending, especially China, where it made up approximately 40% (International Labour Organisation 2011). Debates on fiscal stimulus typically assume government investment is impressive. For example, Paul Krugman wrote in the brand new York Times (Krugman 2016):

“[T]he simplest, most reliable response to a downturn will be fiscal stimulus – preferably government shelling out for much-needed infrastructure.”

I have already been studying the fiscal multiplier connected with government investment, and comparing it to the multiplier of government consumption spending (Boehm 2019). The fiscal multiplier measures by just how many dollars GDP increases after a one dollar upsurge in expenditure.

I found a large and conventional class of macroeconomic models predicts that the federal government investment multiplier for short-lived shocks (like the stimulus through the Great Recession) is small – typically below 20 cents on the dollar. This contrasts with the federal government consumption multiplier, which is between 0.6 and 1 under standard assumptions on parameters.

THEREFORE I estimated government consumption and investment multipliers in a panel of OECD countries. The info broadly support the theory’s predictions: The estimated government consumption multiplier is just about 0.8, and the federal government investment multiplier near zero.

Why could the federal government investment multiplier be small?

Standard macroeconomic models predict that the investment multiplier is small because private investment falls drastically after temporary government investment shocks – referred to as ‘crowding out’. Intuitively, a transitory upsurge in government purchases raises the cost of goods today in accordance with tomorrow. This prospect of falling prices leads households and firms to delay expenditures, offsetting the federal government upsurge in demand.

The amount to which purchases could be delayed is a lot higher for durable investment goods such as for example cars, manufacturing equipment, or structures, than for nondurable consumption goods such as for example groceries or haircuts (e.g. Mankiw 1985). Consequently, a big class of models predicts that the federal government investment multiplier connected with transitory expenditure increases is smaller compared to the government consumption multiplier.

Evidence for a little government investment multiplier

To estimate government consumption and investment multipliers, I build on prior work by Blanchard and Perotti (2002), Ramey (2011), and Auerbach and Gorodnichenko (2012). Figure 1 illustrates the estimated responses after a one-dollar upsurge in government expenditures.

Figure 1 Estimated GDP response to improve in government expenditure on consumption and investment, and crowding out

In response to a government consumption shock, GDP rises above zero and becomes statistically significant in the next quarter (Panel A). On the other hand, GDP remains indistinguishable from zero at all horizons after a government investment shock (Panel B). The associated multipliers are approximately 0.8 for government consumption, and near zero for government investment.

Panel C demonstrates that private consumption remains roughly unchanged after a government consumption shock, suggesting that there surely is no offsetting drop in private demand. At no horizon may be the response significantly not the same as zero at the 10% level. This contrasts with the behaviour of private investment shown in Panel D. After a government investment shock, private investment falls significantly below zero – with out a lag. The estimates become insignificant in the sixth quarter, but remain several standard error below zero before eighth quarter. Hence, the info support the theories’ prediction that private investment is crowded out, and the federal government investment multiplier small.

Recent work has argued that the federal government expenditure multiplier could possibly be greater when the monetary authority is constrained by the zero lower bound on nominal interest levels (for instance Christiano et al. 2011), or when factors of production are underutilised as may be the case during downturns (Auerbach and Gorodnichenko 2012). When I tested these predictions separately for government consumption and investment, I came across some evidence for larger multipliers at the zero lower bound. No robust differences emerge when I estimate separate multipliers for alternative states of the business enterprise cycle.

Multipliers that mislead

A little multiplier for temporary increases in government investment will not imply government investment is undesirable. Actually, research shows that government capital is productive (Aschauer 1989) and discovered that the federal government investment multiplier could be large over time (Baxter and King 1993). Because the productivity ramifications of government capital operate at a lot longer time horizons compared to the short-lived demand effects connected with fiscal stimulus programs, the choice mechanisms are mutually consistent and evidence towards you need to not be interpreted as evidence against the other.

Yet, my findings raise concerns about the potency of fiscal stimulus targeted on infrastructure. Much like a great many other stimulus programs, the American Recovery and Reinvestment Act contained provisions to improve shelling out for highly durable goods such as for example highway infrastructure, high-speed rail corridors, railroads, airports, and broadband. This investment could be desirable for most reasons, but it isn’t clear that it stimulates aggregate demand. The info suggests that any upsurge in government demand is offset by shortfalls in private demand.

My work also offers implications for the empirical research on fiscal policy. If different the different parts of government expenditures are connected with different multipliers, the multiplier for total government purchases, which is often estimated, may have external validity problems. Building on earlier work by Kraay (2012), I possibly could show that the multiplier for total purchases is approximately a weighted average of the federal government consumption and investment multipliers, where the weights reflect the composition of purchases.

Since this composition generally differs across applications, the multiplier for total purchases may differ, even if the multipliers of government consumption and investment remain unchanged. For example, if a stimulus program includes a different composition of government consumption and investment to the identifying variation for the estimated multiplier, then this estimate is a misleading guide for policymakers. Similarly, we’d not be expectant of that estimates for the multiplier of total purchases are necessarily comparable across samples and identification schemes.


Aschauer, D A (1989), “Is public expenditure productive?”, Journal of Monetary Economics 23(2): 177-200.

Auerbach, A J and Y Gorodnichenko (2012), “Measuring the Output Responses to Fiscal Policy”, American Economic Journal: Economic Policy 4(2): 1-27.

Baxter, M and R G King (1993), “Fiscal Policy generally Equilibrium”, American Economic Review 83(3): 315-34.

Blanchard, O and R Perotti (2002), “An Empirical Characterization of the Dynamic Ramifications of Changes in Government Spending and Taxes on Output”, The Quarterly Journal of Economics 117(4): 1329-1368.

Boehm, C (2019), “Government Consumption and Investment: Does the Composition of Purchases Affect the Multiplier?” Forthcoming in Journal of Monetary Economics.

Christiano, L, M Eichenbaum, and S Rebelo (2011), “When May be the Government Spending Multiplier Large?”, Journal of Political Economy 119(1): 78-121.

International Labour Organization (2011), “An assessment of global fiscal stimulus”, European Commission-International Institute for Labour Studies EC-IILS joint discussion paper series no. 5.

Mankiw, N G (1985), “Consumer Durables and the true INTEREST”, The Overview of Economics and Statistics 67(3): 353-62.

Kraay, A (2012), “What size may be the Government Spending Multiplier? Evidence from World Bank Lending”, The Quarterly Journal of Economics 127(2): 829-887.

Krugman, P (2016), “A Pause That Distresses”, THE BRAND NEW York Times, 6 June.

Ramey, V A (2011), “Identifying Government Spending Shocks: It’s all in the Timing”, The Quarterly Journal of Economics 126(1): 1-50.


[1] These details is extracted from The American Recovery and Reinvestment Act website. Because the website is currently offline, I accessed a cached version (dated 4 January 2014) at THE WEB Archive.

Leave a Reply

Your email address will not be published. Required fields are marked *