The fiscal-stimulus plan now being considered by Congress is one of the main hopes for the economy's future. Fiscal stimulus on its own will not heal the economy—the financial system must be nursed back to health, and that will likely require hundreds of billions of dollars more in public funds than have already been committed—but it is a key element in any recovery plan. A massive hole in demand is emerging as consumers, businesses, and state and local governments are forced to cut back. The federal government is the only entity that can fill that gap, either by spending itself or by providing the financing for spending in the rest of the economy.
We are not yet ready to publish a full evaluation of the stimulus plan that is working its way through Congress. But we can look at the impact of the plan that we incorporated in our January forecast, and give some initial thoughts on how the final plan will differ from that.
What Did We Assume in January?
IHS Global Insight's January baseline forecast contained a stimulus package, which was our best estimate as of the end of December of the package likely to be enacted. We assumed spending and tax cuts totaling $500 billion as of the end of 2010 and $666 billion as of the end of 2011.
We assumed three major types of stimulus. First, support for consumer spending through personal tax cuts and transfer payments (e.g., more generous unemployment insurance benefits). Second, increased transfers to state and local governments, to prevent draconian cutbacks in state and local spending. Third, increased infrastructure spending (also channeled through the states).
The following table summarizes the package assumed in the January baseline both year-by-year and cumulatively for major categories of taxes and spending. The figures refer to funds actually spent—not merely appropriated.
Fiscal Stimulus Assumed in January 2009 IHS Global Insight Forecast
(Billions of dollars, calendar years)
Personal Tax Cuts/Transfers
State and Local Transfers
Personal Tax Cuts/Transfers
State and Local Transfers
* In this table we define tax cuts as in most public discussions of the stimulus plan.
To the extent that these tax cuts exceed tax liabilities, they will partly
be classified as transfers in the federal accounts.
What Difference Did Fiscal Stimulus Make to the January Forecast?
What would the economy look like without fiscal stimulus? To address this, we have produced an alternative scenario based on our January 2009 forecast that removes the stimulus. The lack of stimulus means that the collapse in private spending drives the economy down much further. Without the stimulus, GDP growth stays negative for all four quarters of 2009, and the year averages minus 3.6% growth, instead of minus 2.5%. And the recovery in 2010 is far more anemic, with growth of just 0.7%, instead of 2.2% in the baseline.
Without the stimulus, the unemployment rate rises to 10.2% in mid-2010, a full percentage point above the baseline. The loss of jobs is both deeper and more prolonged. Without stimulus, the cumulative loss in jobs peaks at 6.9 million in the second quarter of 2010, rather than at 5.0 million in the fourth quarter of 2009.
What Are the Multipliers for Fiscal Stimulus?
We find that the spending components of the fiscal stimulus have the largest "bang-for-the-buck" (extra GDP delivered per dollar of stimulus). The following table shows fiscal multipliers calculated from our simulation (broken down by the major components of the stimulus). We have focused on the short run and have calculated the multipliers as the average boost to GDP over the first two years of the stimulus divided by the average spending increase or tax cut over that period.
The spending components deliver multipliers in excess of one, while the multiplier for the personal tax cuts and transfers is less than one.
(From comparison of baseline and "no stimulus" simulations)
Personal Tax Cuts/Transfers
State and Local Transfers
Our finding is that the most effective component of the fiscal stimulus is the infrastructure spending (multiplier 1.7). This should not be surprising, since the spending creates GDP both directly (by putting idle resources to work) and indirectly (since those businesses and workers receiving extra income will then be able to spend more).
The transfers to state and local governments are also highly effective (multiplier 1.4). It is hard to describe them as boosting the economy—the point here is simply to prevent a more severe downturn—but when compared with a scenario without them, they give powerful support. State and local workers, for example, who retain jobs that they otherwise would have lost, will have more income to spend than otherwise.
The personal tax cuts/transfer payments have the smallest bang-for-the-buck (multiplier 0.6). They have no direct effect on GDP, and boost activity only when spent. Some portion of the tax cuts will be saved, despite the pressure on family budgets. Households are trying to rebuild their financial assets, and they are also likely to suspect (rightly) that taxes will be going up in the long term. In addition, personal consumption has a relatively high import content, so there is a greater leakage of demand overseas.
Two caveats to these results should be borne in mind. First, these results apply only to an economy with plenty of spare capacity, so that extra spending does not crowd out private activity. Under current circumstances, that is not a problematic assumption. Second, the results do not necessarily imply that the GDP directly created by the government spending is of value. If the infrastructure spending goes for "bridges to nowhere," then the national accounting assumption that the value of government spending on goods and services is represented by its cost will grossly overstate its true worth. All will depend on the projects chosen.
How Do Our Results Compare with the Obama Team's Findings?
The incoming Obama economics team has published a report in which they try to estimate the impacts of stimulus (The Job Impact of the American Recovery and Reinvestment Plan, Christina Romer and Jared Bernstein, January 9, 2009). How do our results compare with theirs?
We know that Romer and Bernstein assumed a stimulus plan of around $775 billion, although they acknowledge that not all of that stimulus would be injected in the first two years. Their overall stimulus is probably larger than ours, but we do not know its breakdown. The most striking comparisons are:
- They show roughly the same relationship between the GDP boost and the employment boost that we do. They use a rule of thumb that says that 1% extra GDP corresponds to 1-million extra jobs; our model shows a very similar relationship.
- Their "no-stimulus" case is more optimistic than ours. They show the unemployment rate rising to a peak of around 9% in the first half of 2010, more than a percentage point below our projected rate. As of the fourth quarter of 2010, they show a real GDP level in the no-stimulus case that is 2.5% above ours.
- And they show a bigger positive impact on output and employment from the stimulus than we do (most likely because they have assumed a larger and more-quickly injected package). They show a stimulus boost of 3.7% to the level of GDP and 3.7 million to employment as of the fourth quarter of 2010; we show a 2.5% GDP boost and a 2.6-million employment boost.
How Do the Plans Going Through Congress Differ from Our January Assumption? Will They Be More Effective?
The packages now under consideration in the House and Senate total around $825 billion. On the face of it, that looks far bigger than our package, since we had assumed $500 billion over the first two calendar years and $666 billion over the first three. Bear in mind, though, that the $825 billion refers to funds appropriated, not funds spent. Much of the spending authorized will only take place over many years, stretching into the next decade, as spending plans take time to be formulated, approved, and executed.
The Congressional Budget Office prepared an initial analysis of the House stimulus plan that reviewed $355 billion of discretionary spending authorized under that plan. It concluded that just $136 billion of that spending would actually take place in the first two fiscal years. So that would effectively cut down the $825 billion plan to $606 billion in stimulus over the first two years. The administration has declared its intent to spend the funds more quickly than that, but it remains the case that the actual stimulus injected over two years will fall well short of the $825 billion headline.
That said, the congressional plans do have an extra element that we had not included, and one which will clearly deliver extra stimulus over the near term, in their provisions for business tax cuts. The Joint Committee on Taxation's markup of the plan being considered this week by the Senate includes $115 billion in business tax cuts over fiscal years 2009 and 2010. The largest components of this stimulus are five-year carrybacks of 2008 and 2009 net operating losses ($59 billion) and extended bonus depreciation ($39 billion).
These business provisions will put cash into companies' hands—but whether they will act as effective stimulus is questionable. They mean more tax write-offs up front—but these will be mostly offset by fewer write-offs in 2011 and beyond. So, the $115 billion of tax cuts in 2010 and 2011 is followed by a total of $73 billion in tax increases over the next four years (2011–14).and another $18 billion over the following five years (2015–19). The net business tax cut over the 2008–19 period amounts to $25 billion.
The bulk of the business tax cuts should be viewed as interest-free loans from the government. Their marginal effect on investment and hiring is likely to be limited, because companies will probably be more interested in using the extra cash flow to shore up their balance sheets rather than boost their capital spending or hiring. So, the multiplier for the business tax cuts is likely to be lower than for the other components of stimulus that we have already included.
There is perhaps just as much risk that the stimulus will be less effective than we have assumed. We assume that after about four quarters, 50–60% of personal tax cuts are spent. But in present circumstances, households are trying to rebuild their financial wealth, so the spending propensity could be lower. And households will also be aware that the size of current budget deficits, combined with the pressure from future obligations for Social Security, Medicare, and Medicaid, means that sooner or later taxes will be going up.
Our longer-term forecast incorporates sharply higher taxes, as illustrated by our assumption about the effective rate of federal income tax (federal income tax payments divided by our proxy for the federal income tax base). Of course, the tax hikes may not necessarily all come via income taxes. Other sources of revenue could be found—the notion of a value-added tax will probably be revived again at some point. But the key point is that the tax burden will eventually rise, and not just on the "rich," if the budget deficit is to be brought down. The more widely this prospect is recognized, the less likely are consumers to spend freely in response to their stimulus tax cuts.
There is also the question of delay. We had assumed that the plan would be ready close to Inauguration Day. It now looks like it will not be ready before mid-February, at the earliest. A delay of a few weeks may not seem like much, but in present circumstances—with the economy in free-fall in the fourth quarter and (we believe) starting off 2009 the same way—time is precious.
The Bottom Line
- We are optimistic that the fiscal-stimulus package will make a substantial difference to the outlook.
- We do not expect that it can give as much help to the economy as the Obama team hopes, even assuming a large dollar-value package than the one in our baseline. We are doubtful about the stimulative impact of the business tax cuts.
- But we believe that the stimulus package can keep the employment "floor" about two million higher than it otherwise would have been, and prevent the unemployment rate from breaching 10%.
- Spending has a high multiplier, but takes a long time to organize. Tax cuts have a smaller multiplier, but can be executed more quickly. These constraints make it difficult to quickly inject an extremely large stimulus, and underline that stimulus can do more for 2010 than for 2009. With every day that passes, the potential impact on 2009 is diminishing.