To overcome the recession following the financial crisis of 2007-09, US authorities have pursued ambitious policy measures aimed at maintaining the liquidity and solvency of the banking sector to ensure a steady flow of credit from financial markets to businesses. Despite these policy efforts business investment declined enormously in the aftermath of the financial meltdown as depicted in Figure 1.
Figure 1: Aggregate accumulation rate for the US non-farm non-financial business sector (Source: Fed)
Why did investment decline? One view is that firms faced credit constraints and were not able to borrow on credit markets to finance their desired investment. Another view is that firms were not willing to invest and borrow due to large idle capacities and low business prospects and were therefore demand constrained. For economic policy the question of which constraint is binding is crucial, as the policy prescriptions differ substantially.
We investigate these constraints by means of panel econometrics to study the determinants of firm-level investment and how they changed over the course of the crisis. For the recent recession, we find support for the latter view rather than the former. Firms did not seem to face finance constraints but were not willing to invest due to low utilization of existing capacity and low expectations. It seems that US authorities were successful in keeping the banking sector liquid. According to our results, however, means to stabilize demand and improve business prospects have not been utilized sufficiently.
What do the determinants of investment tell us about credit and demand constraints?
Assuming imperfect capital markets due to the presence of incomplete information between lenders and borrowers, a various determinants of investment can be justified theoretically (cf. Fazzari et al. 1988 and Bernanke and Gertler 1989). First, the firm’s cash flow is typically used to measure the firm’s constraints in raising external funds. If the firm does not face borrowing constraints then the cash flow should be irrelevant for explaining investment as firms can borrow whatever needed to finance the desired investment. Yet, if a firm does not have access to external borrowing, it can only invest the funds raised internally and investment will depend on cash flow to a great extent.
Second, expected profitability is typically approximated by the ratio of capital stock’s market value to its replacement cost (Tobin’s q) as well as sales growth additionally capturing the effect of idle capacity on investment decisions.
Now, credit-constrained and demand-constrained investment constellations can be easily characterized. The former implies investment to respond sensitively to changes in cash flow but not to changes in expectations and utilization as captured by Tobin’s q and sales growth. Equivalently, demand-constrained investment features high demand and low cash-flow elasticities. A rising cash flow does not cause investment to expand considerably as the binding constraint for firms is demand expectations. If expectations improve, investment expands as it is not constrained by access to external funds.
Empirical investment regimes
To study the constraints to investment historically we estimate the credit and demand elasticities of investment over time. Hence, a dynamic linear investment function is specified and estimated for a panel of US corporate businesses. We use quarterly firm-level data ranging from 1975Q1 to 2010Q4 obtained from S&P’s Compustat North America Fundamentals Quarterly database.
Since we are interested in the cyclical behavior of the coefficients, we apply a rolling window estimation procedure in order to obtain time-varying coefficients. We recursively estimate the investment function for samples spanning over five consecutive quarters and moving from the beginning to the end of the period considered.
We can use the scatter plot depicted in Figure 2 as guidance to identifying the constraints to investment. The figure plots the demand effect against the cash-flow effect for each quarter with the color intensity indicating the state of the economy.
Figure 2: Investment regimes in the US: demand effects vs. cash-flow effects
Times of economic depression are concentrated in the eastern quadrants implying that demand had typically large effects on investment. Most of these quarters are located in the south-east quadrant implying demand-constrained investment. Further, it seems that it was mostly times of expansion during which firms were credit constrained. We obtain an unambiguous result for the recent recession: investment was demand constrained with large demand and low cash-flow effects on investment. Only in the early recovery, the economy moved into the credit-constrained quadrant.
The failure to stabilize investment through fiscal spending
How can we make sense of the surprising result that firms tended to be demand constrained rather than finance constrained in the recent period of economic depression? Part of the answer may be found in the relative policy efforts to sustain the functioning of the credit markets and to generate aggregate demand and improve business prospects.
Undoubtedly, US authorities took historically unprecedented measures to stabilize the credit markets. Since the federal funds rate hit the zero lower bound by the end of 2008 and the economy responded too slowly to the declining short-term interest rate, the Fed made excessive use of unconventional monetary policy through existing as well as newly created temporary lending facilities. It seems that these measures were successful in keeping the credit markets liquid.
However, we would argue that the policy efforts to stabilize aggregate demand were insufficient which may have contributed to the decline in private investment given the large demand elasticity of investment during the crisis. The federal funds rate has traditionally been the main instrument used by the FED to stabilize demand. However, this instrument ceased to be at hand by the end of 2008 when it hit the zero lower bound.
To assess the relative fiscal policy effort to stabilize demand, we consider two measures: First, the percentage deviation of actual GDP from potential GDP, i.e. the relative output gap depicted in the first panel of Figure 3. As long as there is no crowding out of private spending through public spending, the relative output gap indicates the extent by which the public sector fails to compensate the downturn of private demand. Since the assumption of no crowding out may be too restrictive we, secondly, consider the ratio between the real primary deficit cumulated over the period of economic stagnation and the real negative output gap as an average from the beginning of the period to the trough, i.e. the maximum negative output gap. This measure is depicted in the second panel in Figure 3. It is fairly robust to crowing-out effects.
Figure 3:Indicators for monetary and fiscal efforts to stabilize aggregate demand
According to the policy measures quantified in Figure 3, the relative fiscal policy effort has been far below trend in the recent economic crisis. The relative fiscal expansion is moderate despite two major stimulus packages launched by the US government. These measures were only small in scope given the speed and extent of the downturn which has been argued by many Keynesian economists and is confirmed by the plots in Figure 3.
It is not implausible to argue that the observed coefficients for the recent economic downturn are the result of the adverse weights given to the stabilization of demand and the stabilization of credit flows. It seems that the policy measures taken were sufficient to eliminate credit constraints but insufficient to stabilize investment through raising aggregate demand and re-establishing business confidence. Hence the decline of investment starting in 2008 is due to a lack of demand management rather than attempts to maintain the access to credit.
According to our findings, severe downturns of business investment which occurred in the contexts of the recessions in 1982, 1990, 2001 and 2008/09 typically follow a similar pattern. In the early stage until the recovery, investment has often been demand constrained rather than credit constrained. Especially during the recent economic crisis investment responded very sensitively to changes in business prospects and capacity utilization and rather insensitively to changes in available internal funds. In the recovery, however, when expectations improve and demand rises firms increasingly hit credit constraints. This also holds for the recent recovery.
This phenomenon may be rationalized by the typical policy response to an economic downturn. Credit-flow maintaining measures are quickly taken, as they are mainly conducted by the monetary authority. Supporting the banking sector by bail-outs, purchasing of troubled assets and guaranteeing debt by the public does not face strong political resistance. Hence, policy efforts aimed at maintaining the liquidity of the financial sector are usually timely and large in scale. In contrast to that, conducting demand management is more difficult in practice. Interest rate policy involves some delay until transmission effects appear. Deficit spending faces strong political resistance and is therefore usually late and small in scale.
Bernanke, B. S., Gertler, M. (1989): Agency Costs, Net Worth and Business Fluctuations, The American Economic Review, 79(1), pp. 14–31.
Fazzari, S. M., Hubbard, R. G., Petersen, B. C. (1988): Financing Constraints and Corporate Investment, Brookings Papers on Economic Activity, 1988(1), pp. 141–206.
Schoder, C. (2013): Credit vs. demand constraints: the determinants of US fixed investment over the business cycles from 1977 to 2011, IMK Working Paper 106, Macroeconomic Policy Institute, Düsseldorf.
 The heat map indicates the business cycle. The more intense the color, the poorer is the state of the economy.
 The demand effect is the average of the effect of sales growth and the adjusted effect of Tobin’s q. Since a considerable share of each of the time-varying coefficients is determined by factors independent of the business cycle, we use for the construction of the relative demand and cash-flow effects the cyclical components of the coefficients by applying an HP-filter (λ=1,600) to the unfiltered coefficients estimated.
 Note that the recession in the late 1970s/early 1980s is very peculiar in the sense that it featured both large demand and cash flow effects. This may indicate that policy options have not been utilized sufficiently.
 Among other measures, the Fed extended its balance sheet to raise the monetary base and provide liquidity (quantitative easing), growing from $869 billion in August 2008 to $2,882 billion in July 2011. The central bank also shifted its portfolio towards riskier assets (qualitative easing) to take out risk from financial institutions’ balance sheets to foster private lending.
 Taking into account the fact that the potential output has been revised downwards considerably after 2007, the figures for the current recession overstate the fiscal policy effort.
 The Economic Stimulus Act of 2008 involved $168 billion in tax rebates for consumers and businesses. The second major stimulus package was implemented by the $787 billion American Recovery and Reinvestment Act of 2009 which mainly included tax reliefs, funding for states and public investment in infrastructure, health and education.