Urging Congress to act quickly, U.S. Rep. Lloyd Doggett recently laid out his case for renewing "emergency" extended unemployment insurance for another year.
In a Dec. 8, 2011, opinion piece on the Huffington Post website, the Austin Democrat quoted a little "Workin’ Man" from the Nitty Gritty Dirt Band and then listed bold-faced "Facts" including, "Fact: One economist estimates that for every $1 we spent on unemployment insurance benefits, we get $1.61 in economic activity back."
That rang a bell. We were already checking a nearly identical statement on a Web page posted by the Democrats of the U.S. House Ways and Means Committee, and we’d found that another PolitiFact writer checked a similar claim in 2010.
Were they all the same? Basically, yes, Doggett spokeswoman Sarah Dohl told us: All three statements relied on a single estimate from Moody’s Analytics chief economist Mark Zandi.
The factoid was used, then as now, as an argument for extending the federal Emergency Unemployment Compensation program, created in 2008 and originally set to expire in a year. It’s been extended eight times, and the current law providing support for up to 99 weeks expires Jan. 3, 2012. If it isn’t extended, most people who lose their jobs after that will get a maximum 26 weeks of state unemployment benefits.
During a July 2010 Congressional standoff over renewing the benefits, PolitiFact rated U.S. Sen. Jeanne Shaheen’s version of the statement Half True. She got Zandi’s number about right, PolitiFact observed, but other experts disputed the economist’s results.
Dohl told us the $1.61 estimate came from Zandi’s testimony to the House budget committee on July 1, 2010.
Next, we checked with Zandi to see if his $1.61 number still held. He told us that he updates such numbers every quarter and that his most recent figure, for the second quarter of 2011, is $1.55.
Why the decrease? Zandi said new data are confirming suspicions that extending unemployment benefits means some workers just stay unemployed longer. "If they are not working, then less is produced and the economy is smaller," he said.
This "disincentive" effect includes some fraud, Zandi pointed out. But it isn’t always dishonest -- sometimes the worker uses the time to find a better job, and if that worker is an unemployed rocket scientist who could earn $125,000 a year, it’s much better for the economy if he or she is earning and spending that cash instead of living on the minimum wage.
We also asked Zandi how he generates such estimates in the first place. How might handing $1 to an unemployed worker make the nation’s gross domestic product eventually increase by $1.55?
The basic idea is that Uncle Sam hands an unemployed person $1, and that person buys, perhaps, groceries with it; the grocery owner in turn will need to order more vegetables, and the dollar continues to change hands, losing a few cents in some transactions, perhaps gaining in others. Some value leaches away into transaction costs, taxes or savings; but the dollar can generate growth, too -- for example, if the grocery store owner has to hire more staff.
Zandi said he uses a complex econometric model with hundreds of equations to determine whether a dollar spent on unemployment compensation drives the GDP up or down and how much. Essentially, he told us, you take a certain period of time, figure out how much was spent on unemployment insurance and how much GDP went up, and try to hold all the other factors equal.
He also produces estimates on how much bang for the buck the U.S. government gets from many other spending options. That’s part of what he was telling the House committee: which policy options he believes can boost the economy most.
A few examples from Zandi’s Aug. 26, 2011 report on the economy:
• Spending $1 on unemployment insurance boosts GDP by $1.55.
• Cutting corporate tax rates $1 will boost GDP by 32 cents.
• Cutting taxes $1 across the board will boost GDP by 98 cents.
• Spending $1 to increase food stamps temporarily will boost GDP by $1.71.
In its most ridiculously oversimplified form, the equation to get that first number could be written as $1 times X equals $1.55. "X" is the multiplier -- and if the multiplier is larger than 1, your money goes up. If the multiplier is less than 1, your money goes down.
So in this chart, Zandi’s stating that the multiplier for corporate tax rate cuts was 0.32; the multiplier for across-the-board tax cuts was 0.98; and the multiplier for increasing food stamps was 1.71.
Doggett’s spokeswoman had also sent us a study, commissioned by the U.S. Department of Labor and others, that used Moody’s models (and some of Zandi’s work) to conclude that during the recession that began in 2008, every dollar spent on unemployment insurance boosted the gross domestic product $2.
Different results are generated by different time periods, different equations and different data sets, as with the Labor study, but Zandi said the main thing to note here is that each multiplier is not a fixed point. "The multipliers depend on the state of the economy. If the economy is operating full-out, then the multiplier will be low and could even be negative. If the economy is very weak, then the multiplier will be large," he told us.
But herein lies the foundation for one objection to Zandi’s models: Some economists believe it’s not realistic to predict any GDP increase.
Harvard economist Robert Barro is prominent among them, writing Aug. 24, 2011 in the Wall Street Journal that "this idea — that one can magically get back more than one puts in — conflicts with what I will call ‘regular economics.’"
So there’s one fundamental objection. PolitiFact reported more concerns last year, including:
• An economist from the libertarian-leaning Cato Institute cited a survey of 10 scholars’ research -- using different methods -- indicating that $1 spent in tax cuts might increase GDP, but $1 in increased government spending would decrease GDP. That is, for tax cuts the multiplier is above 1, but for government spending the multiplier is below 1.
• Some researchers say unemployment spending will drag the economy down long-term, because the government will eventually have to raise taxes in order to keep paying out.
• Scholars from the conservative Heritage Foundation said it’s invalid to assume the unemployed person will rush out and spend the whole dollar.
Zandi agreed that if the unemployment benefits are "financed by a larger deficit, then it will result in weaker long-term growth." But, he said, if it "helps avoid a more severe downturn in the economy now, then it could very well be worth the long-term cost."
As to spending the whole dollar, Zandi said his view remains that "most unemployed workers will spend their unemployment insurance very soon after they receive it." If you give a dollar to Bill Gates, he said, the billionaire might not spend it at once, but "many people are living paycheck to paycheck, not thinking 10 years down the road."
So, clearly, economists diverge on whether, and how much, spending on unemployment benefits might boost the economy.
For more perspective, PolitiFact looked at estimates from the nonpartisan Congressional Budget Office. In January 2010, the CBO estimated a dollar in unemployment aid would increase GDP between $0.70 to $1.90 -- so Shaheen’s $1.60 was at the high end, but still in the ballpark.
A Nov. 15, 2011, CBO update says that a dollar in unemployment aid would increase the GDP between $0.40 to $1.90. Zandi’s latest estimate of a $1.55 increase in GDP falls toward the high end of that range.
But that’s not the amount Doggett cited. He’s repeating $1.61, which makes the outcome look rosier than it might really be -- even setting aside other economists’ concerns with making such a projection.
Doggett cites an economist’s projection that the country gets a 61 percent return on each additional dollar in unemployment aid. But his "fact" overlooks another fact: Economists are divided over whether unemployment aid increases GDP. Also, Doggett’s statement specifies a figure from mid-2010 that’s been out of date for about a year. A failure to update it is like failing to update the national employment rates.
The only certain fact here is that an economist did once estimate $1.61.
Our sense is you can’t take Doggett’s statement to the bank. We rate it Mostly False.