Exploring Crowding Out

8. The Keynesian multiplier in the simplest textbook models is 1/(1-mpc). A quick look at data from the United States suggests that the mpc is quite high, at least .9. If people are given an extra dollar, on average they will spend at least 90 cents of it.

a. If we put .9 in the formula for the simple Keynesian multiplier, what do we get?

b. Suppose that government uses fiscal policy to increase spending. It spends another $1 billion dollars. In equilibrium, the model predicts that total income will increase by $ billion due to the increased government spending of $1 billion and induced consumption spending of $ billion.

c. Most economists think the multiplier is much smaller than that implied by the simple formula. One popular principles book by Hall and Liebenstein says that in the short run the multiplier is about 1.5, and in the long run it is zero.1 This implies that in the short run the billion dollars of fiscal policy will increase total income by $ billion due to the increased government spending of $1 billion and induced consumption spending of $ billion.

d. Dividing induced spending in part b by induced spending in part c, we conclude that the simple model if off by a factor of about in predicting induced spending.

e. How seriously do you think we should take a model that makes predictions that are off by this factor?

9. Both the Barack Obama and the George W Bush Administrations tried to use fiscal policy to stimulate the economy. The Bush Administration proposed the Economic Stimulus Act of 2008 to head off recession. It offered tax rebates and was passed in February of 2008. The Obama Administration differed from the Bush Administration primarily in stressing the spending side of fiscal policy rather than the taxing side. Its response to the sharp downturn in economic activity following the collapse of Lehman Brothers in September of 2008 was the American Recovery and Reinvestment Act of 2009. This act consisted of a mishmash of various spending projects and transfers, with a few changes in the tax code thrown in. It became law in February of 2009.

The economic advisors to the Obama Administration tried to predict what the economy would do with and without their stimulus package. Using an the assumption that the average multiplier for the various kinds of spending is 1.6, they produced the graph below.

Project Path of Unemployment

The economy followed neither of the tracks in the graph above. The unemployment rate went higher than either path and declined much more slowly than forecast. (See, for example, here). There are at least three possible explanations for why the forecast was so wrong. It may be that fiscal policy does not work. The multiplier may be even smaller than the Obama advisors assumed. Alternatively, it may be that the economic advisors seriously underestimated how serious the recession would be, and that the stimulus package worked because it prevented a much more serious recession from occurring. Finally, some have argued that you do not see an effect because the rise in federal government spending was partially offset by (and may have contributed to) a decline in state and local government spending.

a. The argument of the text that fiscal policy has four shortcomings—lags, too many side effects, the political nature of the decision making process, and crowding out—was written before 2009. How well does it fare when applied to the American Recovery and Reinvestment Act of 2009?
b. Should we take seriously the argument that, even though the forecast was wildly wrong, the policy worked because otherwise the recession would have been far worse? Can't this kind of argument be made for any policy that goes wrong?

Review Question back Quiz

1. Macroeconomics: Principles and Applications by Robert E Hall and Marc Lieberman, Updated Second Edition (Thompson Southwestern, 2003), p 252: "Most forecasting models used by economists in business and government predict that the multiplier effect takes about nine months to a year to work its way through the economy. At the end of the process, the multiplier has a value of about 1.5."