Economists Finally Find a Cause: Saving ATUS

There is no shortage of groups made up of citizens banding together for a cause: Greenpeace, Doctors without Borders, Save the Children, the KKK, etc. I suspect that if you look at the data, you will see that economists are nearly always underrepresented in these organizations. No doubt there are many factors contributing to this result. In general, economists tend not to be idealistic, and also are often more concerned about their private interests than the public good. Economists also tend to be pretty conservative and not very religious, and my guess is that the lion’s share of activist groups are from the left, or have religious links. Finally, economists tend not to have much faith in governments, so they don’t invest as much effort in trying to spur the government to take action.

Occasionally, large groups of economists do become galvanized behind a cause. Sometimes, that cause is a policy issue, such as free trade. At the present moment, however, the cause is a little known government survey known as the American Time Use Survey (ATUS). Unlike most government surveys, which ask large numbers of people relatively simple and broad questions about their lives, the ATUS takes a small number (currently 14,0000) of people and actually tries to get at what Americans are doing with their time, minute by minute. Thus, these data allow researchers to answer questions that otherwise would be out of reach. For instance, when someone becomes unemployed, what do they do with their time? How much do time do they spend searching for a new job, and how much does that time increase or decrease as their unemployment benefits come closer to running out? Do husbands do any housework? Are black children and white children spending their days doing different activities?

Funding for the ATUS has been eliminated from the proposed Fiscal Year 2009 budget. Many economists are upset because it only costs $6 million to fund the program. Although I could not find hard data on it, I’m certain this survey is far cheaper to administer than others, such as the National Crime Victimization Survey, which, because of restrictions on access to the data, are largely a waste of money. Even though the ATUS has only been around since 2003, it is already used widely by economists. As the number of years of available data increase, it will become more and more valuable.

Katherine Abraham, Suzanne Bianchi, Dan Hamermesh, and Alan Krueger are spearheading the effort to save ATUS which you can read about here.


Did anyone see 'Steven D. Levitt' as a signature on the save-the-atus letter?

Charles D

Economic data is too reliant on the truth for a government to properly back it. Start putting more bias in your research and I'm sure you'll get the money.


There are some things government does well, as well as best.

ATUS is one example of this.

Save ATUS!


Dude, Barack will take care of this.

scott cunningham

Thanks for the headsup! ATUS is a great dataset. I hope this is successful. Hamermesh has a good article explaining why this data is so good here: (


"I suspect that if you look at the data, you will see that economists are nearly always underrepresented in these organizations."
Economists are usually too busy creating the problem in the first place. Russia's clusterfrak of an economy springs to mind, courtesy of Milton Friedman and affiliates.


I'm actually a left-leaning economist and disapprove of the stereotype that we "hate" the government and worship privatization. My stance is to rely on empirical results, and in many cases, the privatization trend has lead to tremendous problems, at least in practice. See Argentina, California energy, etc.

Economists might not be a religious bunch, but I do see a frightening amount of "blind faith" regarding certain old theories within the profession.

I also always found it amusing that those economists that lack faith in governments have no problem using the gov't datasets or asking for such datasets to be subsidized by the public via the gov't (or, in part, by the university-paid subscription fees).

I agree with the above, according to theory, wouldn't a private market develop for such data if the need was high enough? Or are economists just not willing to dig into their own pockets for such things?



Why do I get the sense that such a survey may be prone to error?

It's easy to answer simple, broad questions truthfully. I'm not convinced that polling people on how they spend their unemployed time (at a fine granularity, no less) will enjoy the same level of accuracy.

Who'd have thought? Extrapolating from the results shows that everyone in America dedicates 20 minutes a day to answering surveys!


Maybe if more budget items, even as useful as ATUS, were eliminated-and the related purchasing, receiving and distribution functions minimized the government payroll might decline, and the congress could focus on critical national issues like energy independence and defense.

scott Cunningham

Umut O - Sorry about that. It's in the Journal of Economic Perspectives, vol. 19, issue 1. I think 2005, but not sure. The title is "Data Watch The American Time Use Survey." HEre is another link, though it's gated:

Pradeep Kumar

While supporting the cause to save ATUS, I feel there is a need to make this study more meaningful (and thus more marketable):

ATUS records primary activity only. In today's world, people spend less time in discrete activities.

It is important to identify the extent of simultaneity. Simultaneity becomes even more relevant with the youth and the more digitally involved population.

Providing data in ready-to-use formats for key user groups may require minimal data processing but could make it a lot more attractive and marketable.


How many people would "(currently 14,0000)" be, exactly? If I take my salary and add a couple more zeroes after the comma and without a decimal point, can I claim to be making 6, 7 or 8 figures?

ML Harris

I propose that the economists in question should run their own ATUS and let the government save $6M. I know it's chump change to the deficit, but the road to riches starts with a dime, as the ads used to say. Of course, as government is completely incompetent according to most economists, the economists could run it for a fraction of the $6M. Maybe we should contract the BLS while we're at it. Oh, that would put a lot of economists out of work, so we NEED that data. *sigh* I really liked the book, too.


If ATUS is useful as you say then it should be easy to find a willing private party to pay for it.

Umut O

Hey Scott,

Are you sure about the link? It looks broken. Do you have the name of the paper?



The National Assessment of Educational progress is $90 Million per year and I don't think the test data is nearly important to the public good as the ATUS.

Yet some want to triple the NAEP budget to increase the number of subjects covered, in spite of the fact that there are serious problems with the tests methodology and with the ability to actually use the data (Difficulties Associated with Secondary Analysis of NAEP Data - Sheila Barron, University of Iowa.)

$6 million for usable data vs. $90 for a test used to make political hay...


Freeriders: I don't think that economists will try to save this. The incentive is to let others save it and then use the data from it since the private marginal benefit likely doesn't exceed the private marginal cost of lobbying.


Yeah, neither Levitt nor Dubner even signed the petition to save the ATUS! How hypocritical!