The Limits of Fiscal Policy

The best case for fiscal policy is during a recession caused by an aggregate demand shock. But even with this best-case scenario, it's still difficult to effect change. An ideal stimulus is timely, targeted, and temporary. What do we mean? Well, an ideal stimulus would quickly hire unemployed workers, putting them to work on projects, which were completed as the economy recovers. This is ideal because an unemployed worker has a low opportunity cost -- so that's the best time to hire. If instead, the government hires a worker who would have had a job anyway, the cost to society is much greater.

Let's look at each of these problems in more detail. First -- timing. New government expenditures can take years to move from dream to reality. First, there's the recognition lag. It takes time to identify a problem, such as a recession. Then, there's a legislative lag. Both houses of Congress and the President must approve any government expenditures -- think committee meetings, debates, drafts, complicated language, more debates, budget cycles. This lag alone could take months or even years.

And once government expenditure has cleared Washington, we've only just begun! We haven't even implemented the project yet. This can involve further hurdles at the state and local level, such as selecting a firm and writing a contract to perform the work. It'd be nice if there were a lot of shovel-ready projects, but the reality is that every project requires planning, permitting, environmental review, and other delays. And then, there's an effectiveness lag. It takes time for government spending to ripple through the economy. Wages, for example -- they aren't paid upfront on day one, but only over time. And then it takes time for the wages to be spent, and so forth. By the time that government spending is working its way through the economy, the situation may well have changed.

One way to minimize lag time is to focus on automatic stabilizers -- fiscal policy that occurs automatically, without legislation. For example, if the economy is doing poorly, income, capital gains, and corporate profits fall. Our progressive tax code could even mean lower tax rates during tough times. And that's good because lower taxes will result in more spending. Welfare and unemployment insurance also automatically kick in when individuals are hurting financially. So, these payments, they're also timely and targeted -- targeted not only on the population that needs the funds the most, but on the population that is most likely to spend the money quickly, boosting the rest of the economy. Most economists, therefore, favor automatic stabilizers.

Now, let's turn to targeting. Ideally, we would want to hire unemployed workers, but that isn't always possible. The very term "shovel-ready" suggests construction workers And in the last recession, lots of construction workers did lose their jobs, but so did retail workers. And the government simply doesn't have a lot of ways to hire waiters and store clerks -- at least not directly. Even among construction workers, there are big differences. A roofer, for example, who had been working home construction -- they might not have the right skills to easily switch over to road construction. As a result, when the government spends money on a new construction project, that may mean hiring workers away from other jobs rather than hiring unemployed workers.

Now eventually, labor demand has still increased, but the less targeted the stimulus, the less timely and effective the stimulus. Finally -- magnitude. Spending money -- it's actually harder than you think. In 2015, US GDP was roughly $18 trillion. And federal spending? It was over $3 trillion. But there isn't that much discretion in the budget. Non-negotiables -- things like Social Security, Medicare, national defense, and interest on the debt -- they account for 65% of the budget. Authentic discretionary spending is less than 20%. And even that portion of the budget is not really all up for grabs.

The 2009 stimulus, the US' largest stimulus since World War II, -- it was roughly $900 billion spread over three to four years. Even at its peak, the stimulus was only about 2% of annual GDP -- not bad, but not that large either. So even assuming that everything else goes right in theory and practice, it may be difficult to spend enough to fully offset an aggregate demand shock. That doesn't mean it isn't worth doing. Most studies of the 2009 stimulus find that it probably did increase GDP and reduce unemployment. In a way, this was an ideal case for fiscal policy precisely because targeting and timeliness are less important when the recession is very severe and lasts a long time.

Let's summarize. An ideal stimulus is timely, targeted, and temporary. But none of this is easy. In a severe recession, stimulus may be extremely valuable. But as with monetary policy, it's a lot more difficult than just shifting lines on a graph.