The 3.5% growth rate was, in my view, in large part attributable to direct measures to stimulate the economy, including direct spending on goods and services by the government (Federal, state and local), as well as tax measures. First, let’s take a look at how each category of final demand accounted for total growth, in the context of a mechanical decomposition, in Figure 1.

Figure 1: GDP growth and contributions to growth of GDP, in ppts; GDP (black), consumpion (red), fixed investment (green), inventory investment (orange), government consumption (purple), and net exports (light brown). Non-shaded area denotes 2009Q3 advance release. Source: BEA, 2009Q3 advance release, October 29, 2009.

Figure 1 breaks down the contributions of overall growth into the broad national income accounting components (overall investment decomposed into fixed and inventory investment). Interestingly, the contribution of government is fairly modest in Q3 (Note change of vertical axis scale).

Figure 2: Government contributions to growth of GDP, in ppts; total government (gray), Federal non-defense(pink), defense (teal), and state and local (brown). Non-shaded area denotes 2009Q3 advance release. Source: BEA, 2009Q3 advance release, October 29, 2009.

Figure 2 breaks down the government consumption growth contribution into that from Federal nondefense, defense, and state and local government.

Government spending on goods and services (not overall government expenditures) accounted for 0.48 percentage points (ppts). Federal nondefense expenditures accounted for 0.17 ppts, while defense accounted for 0.45 ppts. State and local spending accounted for negative 0.14 ppts. At this juncture, one could leap to the conclusion that the stimulus package, and other measures, had no effect on output. And I’m sure many will. But I think it pays to be a bit circumspect in this regard.

First, it’s always helpful to recall that the advance estimate incorporates lots of estimates, and is subject to revisions (see this post).

Second, some individuals have argued that since a portion of the government component comes in the defense category, that should not be construed as being attributable to the stimulus package. But in point of fact, according to CRS ARRA does have some defense expenditures (mostly energy efficiency upgrading). One can see what contracts have been let by going to the http://www.Recovery.gov website (noncompetitive contracts here). As an open question, I’m not sure where Army Corps of Engineers expenditures fall in the categories (I think it’s under defense as well, in which case the defense category would incorporate even more of the stimulus spending).

Third, the decline in state and local government spending’s contribution is notable. Given the big budget shortfalls in state budgets [1], what this outcome tells me is in the absence of the transfers from the Federal government, the negative contribution would have been even larger.

The Joint Economic Committee held hearings today; J Steven Landefeld, the head of the BEA, stated in his testimony:

…let me conclude by describing how it is reflected in GDP and the national accounts. BEA’s national accounts include the effects of the federal outlays and tax cuts included in the ARRA. Because most of the outlays and tax reductions from ARRA during the last three quarters were in the form of grants to state and local governments, tax reductions for individuals and businesses, and one-time payments to retirees, their effects on GDP show up indirectly through the effects on GDP components such as consumer spending, residential investment, and state and local government spending. Thus, BEA’s accounts do not directly identify the portion of GDP expenditures that is funded by ARRA. During each of the second and third quarters, the Making Work Pay Credit lowered personal taxes and raised disposable personal income about $50 billion (annual rate). During the second quarter, ARRA provided payments of $250 to beneficiaries of social security and other programs that raised disposable personal income about $55 billion. ARRA also provided special government benefits for unemployment assistance, for student aid, and for nutritional assistance; these special benefits raised disposable income about $49 billion in the third quarter and about $35 billion in the second quarter. ARRA also funded grants (such as Medicaid) and capital grants (such as highway construction) to state and local governments of about $75 billion in the third quarter and $85 billion in the second quarter.

Mark Zandi also testified. His testimony included this interesting table, reporting estimates of expenditures.

Table 2 from Zandi .

Last 5 posts by Menzie Chinn

• China, the Renminbi, and Global Imbalances: A Quantitative View - November 20th, 2009
• GDP: Revisions and Forecasts - November 19th, 2009
• Assessing the Impact of Government Policy on Widget Consumption and Widget Sector Capital Usage - November 16th, 2009
• The Global Surface Temperature Anomaly - November 16th, 2009
• Politico Does Economic Analysis... - November 11th, 2009