Deutsche Bank Blows It With Rosy Payroll Tax Projections

There’s an old adage: you can fool some of the people some of the time, but economists are foolish people most of the time.


Economists at Deutsche Bank prompted us to recall this bit of wisdom today when they raised their projections for GDP growth next year from 3.3 percent to 4.1 percent. Their reason for the raise was that the tax deal agreed in Washington, DC last night promises a payroll tax holiday. Their basic reasoning seems straight forward enough—if you increase the amount of money that shows up in the paychecks of workers, they’ll go out and spend most of that money, which will help the economy grow.

Unfortunately, it’s not so simple. In fact, temporary tax relief tends not to increase consumer spending by very much. What’s more, tax relief that comes in the form of a temporary payroll tax cut is even less likely to stimulate spending.

But first let’s give the Deutsche Bankers their due. As you’ll see, they have lots of math on their side.

Wages and salaries currently total $6.44 trillion, and they have been growing by nearly 5 percent over the past couple of quarters. At that rate, wages and salaries will total about $6.75 trillion at the end of next year.

Presently, individuals and their firms each contribute 6.2 percent up to a cap of $106,800 for Social Security. Under the payroll tax holiday, the individual’s contribution will temporarily fall to 4.2 percent. Based on the Social Security taxes paid in fiscal year 2009, we estimate that roughly 85 percent of total wages and salaries are subject to the payroll tax. This means a 2% reduction on payrolls taxes equates to a $115 billion increase in wage and salary income (i.e., 85 percent of $6.75 trillion is subject to the payroll tax, and a 2% increase in the product equals $115 billion). With the personal savings rate currently at 5.8%, we estimate that $108 billion of the “income gain” resulting from the payroll tax holiday will be spent. This is worth 0.7 percent of 2011 GDP. Consequently, the payroll tax holiday could give the economy an added fillip next year in addition to any incremental benefit from improving financial conditions, the full extension of the Bush-era tax cuts and the expanded business investment tax credit.

(Hat tip: Business Insider)

The silent assumption here is that consumers will spend their increased income from the payroll tax holiday in the same way they spend their regular income. That assumption, however, is way too optimistic. In the first place, as good old Milton Friedman taught us, people tend to spend based on their expectations of future income instead of on the reality of current income. One of the lessons of this is that temporary tax relief doesn’t increase spending by much because people know that their future income will take a hit when the tax comes roaring back.

We’ve seen the effect of this—economists call it the Permanent Income Hypothesis—both before and after the housing bubble burst. While the housing bubble was growing, people had high expectations about future income and wealth. This showed up as a diminishing ‘savings rate’—which meant that the spending of households was growing faster than savings. What really was occurring, however, was the people were deciding they could afford to consume more now on the assumption that they’d be wealthier later. Low interest rates had a hand in encouraging this kind of thinking by creating the illusion that their be more wealth in general in future.

When things fell apart, people reined-in their expectations of future income and wealth. As a result, the savings rate increased. But, again, while it increased dramatically in terms of current income, what really seems to have happened was that people brought their spending and savings more in line with their diminished expectations for the future.

The case against a strong payroll tax holiday stimulus effect goes beyond Milton’s hypothesis or the recent savings rate pendulum swing. In 2004, economists David S. Johnson, Jonathan A. Parker and Nicholas S. Souleles studied the effect of the income tax rebates of 2001. To fight off a post-dot come bust economic recession, the government sent rebate checks, typically ranging from $300 to $600, to households. What the economists found is that people spent only 10 to 40 percent of their rebates in the first three months following the rebate. Over the next three months, the spending increase was much smaller. As a result, it appears as if only two-thirds or so of the rebate resulted in increased spending.

To put this in perspective, the savings rate in 2001 was negative one—that is, people were spending more than they were earning. If people spent their 2001 rebates the way the gang at Deutsche think people will in 2011, they would have spent not just two-thirds of it. They would have spent more than 100 percent of the checks.

Let’s turn this around, then, as see what happens if it turns out that 2011 folks spend their additional income just like people did in 2001. In that case, out of the $115 billion in additional income, only around $76 billion would go into additional spending. That means the gain to GDP wouldn’t be 0.7 percent. It would be less than 0.5 percent.

Even that may be too optimistic. A recent study doneby another trio of economists—Claudia R. Sahm, Matthew D. Shapiro, and Joel Slemrod—looked at how stimulus effects differ depending on how tax breaks are delivered. What they found is that people spend less when they get additional income from payroll tax cuts than when they get a one-time check. This was not some marginal move either—the spending rate on payroll tax cuts was half of what it was for rebates.

What this means is that while people might have spent $76 billion if they got this as a one time rebate check—like they did back in 2001—they will likely spend only half of that—$38 billion—when they get it as an ongoing payroll tax holiday. And just like that $108 billion in stimulus has become $38 billion. For those of you keeping score, that’s a bump of just 0.25 percent or so in GDP.

Why is it that people might retain so much of their additional income, instead of spending it? It may be that savings rates are depressed by fixed-expenses. If you’ve got kids to feed, schools to pay for, car payments, house payments—you may be living close to the edge right now. Even if you wanted to save more, you could not. But given additional income, you’ll be able to come closer to the savings rate you desired in the first place. Higher income, in this case, would be positively correlated with higher savings rates.

To be sure, there is likely to be an economic bump from the tax holiday—and perhaps a welcome one in a time of still anemic growth. But Americans are likely to save an awful lot of the payroll tax holiday income, in a way that may result in financially healthier households. But anyone basing their economic projections—or business plans—on Americans spending their new income just like they did before the tax holiday is likely to be in for a shock.


Questions? Comments? Email us

Follow John on Twitter @

Follow NetNet on Twitter @

Facebook us @