Unemployment benefits and Google job search

I had not known of this Scott R. Baker and Andrey Fradkin paper until recently, here is the abstract:

The large-scale unemployment caused by the Great Recession has necessitated unprecedented increases in the duration of unemployment insurance (UI). While it is clear that the weekly payments are beneficial to recipients, workers receiving benefits have less incentive to engage in job search and accept job offers. We construct a job search activity index based on Google data which provides the first high-frequency, state-specific measure of job search activity. We demonstrate the validity of our measure by benchmarking it against the American Time Use Survey and the comScore Web-User Panel, and also by showing that it varies with hypothesized drivers of search activity. We test for search activity responses to policy shifts and changes in the distribution of unemployment benefit duration. We find that search activity is greater when a claimant’s UI benefits near exhaustion. Furthermore, search activity responses to the passage of bills that increase unemployment benefits duration are negative but short-lived in most specifications. Using daily data, we estimate that an increase by 1% of the population of unemployed receiving additional benefits results in a decrease in aggregate search activity of 1.7% lasting only one week.

One way (not the only way) of reading these results is to wonder if some of the unemployed feel they ought to increase their shirking in response to an extension of benefits, but they actually don’t really want to do so.  They shirk a bit more, for a short while, not to feel like fools, and then return either to active search or fruitless despondent search, as the case may be.  For better or worse, habit dies hard.

For the pointer I thank John Horton.

Conor Sen, by the way, tells us that “ask for a raise” is at a post-recession high on Google Trends.


Someone should test my idea of initially high, but monthly decreasing, unemployment benefits. A relentless message.

Not a bad experiment.

Having spent a brief stint of unemployment twice when benefits were a tiny percentage of my prior income, I would have appreciated a high payment for the two months it took to find an acceptable job. The low payments I got added insult to the injury of having been taxed for it for so many years.

Is your UI tax limited because your benefit is limited, the way your SS contribution is limited because your benefit is limited?

Here's a 35 year old study that considers that idea:


See Table 1 on page 1355.

Not a bad experimental idea at all.

The same should be done for Medicare, If a 95-year-old want's triple-bypass surgery, the program should reimburse a smaller portion than it does for a 65-year-old.

Lots of things that make sense don't make political sense, hence they don't make sense.

In other words, a new nation conceived in liberty will take slightly more than 87 years before croaking. Maybe the Chinese have it right.

Judging by the British experience, it can take a generation or two before the old habits die and you get armies of families who live off the welfare state as a deliberate ambition.

It seems to have taken roughly the same time for the old notions of propriety and responsibility largely to die out among doctors and nurses in the NHS.

Authors do a good job anticipating criticisms of their measure. For those who don't read the paper, I'd point out two things. (1) "Rates of internet usage in job search increased with education but did not vary systematically by census region." Those with more education likely entered unemployment from higher-paying jobs and may thus receive (depending on state) higher benefits. They may also be more likely to hold out rather than take a job that leads to underemployment. Much as in the debate over the minimum wage, we should be careful in our policy discussion to not treat the unemployed as a monolithic population. We might be able to spur the more highly educated into (worse) jobs, but at the cost of sending the lower educated into even worse poverty.

(2) The measure of job search is the search term 'jobs'. It's plausible, and the authors investigated some alternate terms like "temp" and "monster". However, for the sake of the study it is perhaps fortunate that Steve Jobs did not pass away until 2011 (after the data collection period).

Higher benefits for higher paid people are an insult. They are a miniscule fraction of prior income - less than 25% AND taxable.

UI is a criminal redistribution scheme. We should have had fully-funded self insurance a long time ago.

Anyone who tries to point out that UI might be decreasing the incentive to work will immediately be badgered with some anecdotal sob story about a friend of uncle who just can't find work and relies on UI to feed his kids. The person, merely analyzing statistical data, will be labeled a heartless monster who reads too much Ayn Rand and doesn't realize that they were born with a silver spoon in their mouth and just how hard other people have it.

What a pair of certifiable tea-hat fogies.

You ought to go out and try to get a job at your age (anyone much past 40, not to say the rest).

You'll find it's either fast food or telesales, no more than 24.5 hours a week though you'll spend 50 on transit and be required 6 days), back to back midnight and dawn shifts, an $8 per hour, which will require food stamp transfer to be able to eat. Benefits like health or dental? None, so good luck with ever getting sick.

Civilization is what cares for all folk. Recommended.

Oh my goodness! Can you believe that they would actually make unskilled workers do unskilled jobs?! I have no marketable skills, but gosh-dangit I'm old, therefore I deserve a cushy white-collar desk job. I don't care if there's some recent college graduate who has up-to-date technical skills, I have real-world wisdom. It's not my fault I always felt entitled to move up the career ladder regardless of whether I have adapted to changing labor demand.

Worse of all, can you believe that when I start the job I actually have to get entry level hours and benefits! Give those midnight shifts to the 30 year olds, I'm 50! I don't care if they've worked there for 10 years, I deserve the best position. Don't you know my generation gave the world the Rolling Stones!

And that in a nutshell pretty much encapsulates most of our political and economic problems: baby-boomers feel entitled to live comfortable, low-stress lives, materially abundant lives. Regardless of what they have to offer the actual economy. So everyone keeps suffering and sacrificing a little bit more to give the boomers what they demand.

Fifty year olds are on the tail end of the baby boom, it's post-boomers you've got in your sights.

I sympathize, but do keep in mind there are an awful lot of people in every age bracket doing unproductive work and expecting to get paid a lot to do it. It just stands out more when someone has been doing that for decades and so has no idea he/she is not worth that much to the market. It's more devastating to the newly dropped employee, because he's been told for a long time that he's got a certain value. It's distressing for those around because the dropped employee looks like he thinks because he's had a better-than-he-deserved deal for twenty years that entitles him to more of the same.

The answer is not to dress down the laid off old people, it's to open the eyes of people at every level. Those 20 year olds with such dramatically up to date skills? They're usually not as irreplaceable and important as they think they are, either.

Your rant may overall be on-target, but weird hours really really suck. If I have to work four 3-hour shifts, and the boss can call be in for other shifts at will, besides the normal time+money costs of the commuting, it makes it very hard for me to get a second part-time job.

Worse than that, weird hours may be counterproductive for some situations.

I took a job a few years ago for low pay and mileage in a rural area, I wound up busting my car and in the end probably worked about six months and wound up losing money.

I know others in similar situations, the point at which it becomes unreasonable to take a job is not the point at which the wage is zero, particularly if you have to put kids in child care, you wind up paying more for clothing, gas, or food, medical costs, etc. The calculation is going to differ per individual and young people with cheap cars and no family are likely to be able to profit better from low paying or irregularly houred jobs more easily than older people with families. That's not about right or wrong or being willing or not to work, that's just the dollars and sense of taking or not taking work.

That's a good idea and an interesting finding. However, I wouldn't expect EUI to have such a blatant effect on recipient behavior. I would expect job search activities to remain relatively stable. The added financial support would manifest itself more in the selection of jobs to apply for, the reservation wage, etc. These factors would serve to increase the average duration of the typical recipient without changing their gross level of job search activity.

The first few weeks of unemployment are a mix of denial and relief. You enjoy the time off but are still bitter about whatever reasons caused you to be unemployed. Most people see it coming and many actually start their search before unemployment begins. But the first week is a wait week where you get nothing.

My first time on UI, I was in no hurry to find a job: lots of savings, voluntary career change, and single. I spent two months receiving benefits but only casually looking because I was required to make three contacts a week. The counselor didnt understand I no longer wanted to be a lawyer, and didnt understand why I couldnt search for work door to door. Imbecile.

I finally took a job when I got bored watching CSPAN all day.

The second time, I was married so I was more motivated but benefits didnt pay our food bills.

If you had lived in a State that was really stingy on unemployment, maybe even fostering you being considered a contract worker so not eligible for benefits, would you have found a job faster either time?

Good question. Hard to say because in both cases I was well prepared with savings and income from the National Guard.

The fact that I had a professional degree already set my sights somewhat high. I could have possibly found a full time job as a lawyer in minutes if I wanted to continue that. Since I didnt make any phone calls for that, I don't know how well that would have gone.

For high income earners, the benefit payment is almost nothing. From that perspective every payment is stingy. I want to say I was getting about $1200 per month before taxes, replacing a salary of about $8000 per month. It was a government job, and I didn't have severence pay. I used the time off to consider what else I wanted to do in life.

Im not sure it affects our job search at all. If there is a disincentive effect, I would expect to see it either at low wages or in highly specialized blue collar jobs like carpenters. If benefits are 75% of your former salary, you can take a long vacation. If you think a seasonal job will come back, you can also afford to lounge.

From my own experience I can only say that the job search requirement to qualify is minimal and easy to lie about. The economics question isnt how you or I would respond to incentives, but how people in general respond. Thats an empirical issue although analysis of human nature might be insightful.

I do get what you're saying -- you have a mortgage, etc. set on your previous salary.

But you know it can and usually does work the other way, right? If $1200 a month is 75% of your normal salary, you can't afford to take a vacation because that 25% of your salary is needed for essentials, while 25% of an $8000 a month salary ought to be pretty much icing, right?

Granting the fat exceptions of the recent college grad with no attachments and a cheap apartment, or the gal with a husband still earning who wanted to stay home with the new kid anyway, etc.

the job search requirement to qualify is minimal and easy to lie about

A colleague was working in a field where lots of people would come through the door and apply merely so they could get the checkmark to keep on receiving their UI. Sometimes the "candidates" would say it up front. She found it a giant hassle in her workday. For what it's worth, another point in favor of just giving people UI without having to "prove" anything.

One problem is that unemployment has been declining, so they need to adjust for the effect of more job openings (or fewer layoffs) creating, by itself, more of an incentive to job search.

For example, assume that in period 1 there is an 11% unemployment rate, and there are 500 applicants per announced or searchable job.

In period 2, 11 months later, unemployment is 8% and there are 150 applicants per announced or searchable job.

Ask yourself this question: Do the characteristics of each period make it more, or less, likely that you will job search using the internet relative to the other period.

Great point. The data is available from JOLTS, so it can be used to calculate a subjective probability of success.

I'm not sure what the problem is, there seems to be an awful lot of jobs in conducting, supervising, and publishing studies. Can't everybody just do that?

+1. I wonder if they have tenure, and if removing their tenure and exposing them to the pleasure of job search would be an informative personal experience.


Perhaps the solution to the underlying issue is to say that a person's UI has a *lifetime* duration cap of (say) 6 years total. (Limited by how much they've worked of course.)

So rather than saying "everybody gets 99 weeks, including Joe who has had zero weeks of UI in the past 30 years due to Joe's relentless personal effort, and Phred, who has worked just barely maybe 1 hour more than required to be eligible, and at the end of this 99 weeks will have spent more time on UI than actually employed." Maybe we ought to give Joe the full 99 weeks and tell Phred it's 26 weeks and then Pfred's done.

The cap might be adjusted for circumstance - be 6 years total when headline unemployment is above 10% but only 4 years total the rest of the time.

How many people find jobs successfully doing google searches??

I'd like the data that shows that google searches result in being hired at higher rates than, say going out to eat with former coworkers on say a weekly or monthly basis. Or going to weekly job hunting self-help groups. Or going to Rotary or other business groups once a month or maybe once a week.

If random google searches are better than the private job search agencies hired by the States using Federal funds and sometimes local funds, then why are there State employment agencies when free google at libraries would be better.

And what is the justification for Monster and Dice, et al if simple google searches are the best way to get a job.

Or is google the search engine at the heart of all the job offer databases run by Monster et al, the State employment databases, etc so it offers comprehensive data on all job searches everywhere that isn't based on networking?

Still, networking is still the dominant way people actually get jobs, isn't it, especially when employers are flooded with way too many applicants.

Was wondering the same things. Indeed.com followed by headhunters and networking. Google wouldn't be in my top three.

Agreed. I've never known anyone, or known anyone who knew anyone, who got a job from a straight Google search.

Sites designated for certain localities or fields are different.

So I guess this study is not measuring an increase in people trying to find a job so much as measuring an increase in people so desperate to find a job they'll do stuff that is super unlikely to find them a job.

I got my last job from a Google search, around 8 years ago. The one prior to that was the result of a college corkboard posting. Perhaps your lack of evidence came from some other factor common to your peers?

Absolutely maybe, totally anecdotal.

The study is looking at job search behavior. It doesn't matter if few people use it as long as the people who do are representative of the entire population.

The study can be criticized ob many levels, but if the same study were done for Monster.com and by survey with similar results, the finding is bolstered. If results vary, then the next inquiry is whether the people who use different methods have differing characteristics.

I know during the last recession, enforcing search effort as pretty much abandoned in my state, but it was still technically a requirement to receive UI benefits. I wonder how many people actually use web searches to establish the required weekly contacts.

The sad part about most of this discussion about the long-term unemployed is that it seems to be conducted almost entirely by people who have never experienced unemployment for any meaningful period of time.

I was unemployed for 6 months between March and September of 2012. It was one of the most difficult periods of my life. I lost the routine of daily job. I lost the self-respect of earning a paycheck. I had to engage in a real job search every week because my unemployment check required that I contact employers every week. It actually got hard to find jobs to apply for because I had run through most of the available jobs in the area in my desperate and fruitless search for work immediately after having been let go.

I ended up finding a good job not because I found the right employer but because I was lucky enough to have a social network extending back to high school that provided me with a new opportunity. My finding a job had NOTHING to do with my level of effort looking -- I had actually given up more or less and started applying for bullshit jobs that wouldn't use my skills. Unemployment insurance had nothing to do with my thinking. It just helped me to continue paying my bills. I felt like an asshole every time I cashed that check.

+1 I hope others read your comment.

Because your individual experience (and mine) only adds one anecdote to the fire, and nothing we claim here can be verified.

The only thing that matters from a policy perspective is what PEOPLE DO with the incentive structure given. If you want sympathy, call a priest. It's not that we don't care, it's that it jist doesn't matter to this discussion.

By the way, while it is natural to 'feel like an asshole every time you cash that check,' remember that YOU PAID your UI premiums when you were working. No one is saying you are a deadbeat for collecting benefits. As for job search effort, that is individual and particular. How people respond is an empirical issue - looking at data.

The reason UI funds are insolvent is because states didnt run UI as a fully funded program. They ran them as pay as you go. So just when people needed benefits, states didnt have the revenue. Aside from the problem of disincentives for job searches, these benefits ravage state finances. That is part of the policy discussion.

The only remaining issue was whether you paid sufficient premiums based on the risk of future unemployment. That is an actuarial issue.

Shorter conservatives to the unemployed: drop dead

You're right. Depending on where you're located, it may not take long to run through available postings that fit (even remotely) your field - and after that it's just depressing.

Worse, when jobs you want and applied for are reposted and you never got a call for the first one. Makes you think someone in HR is throwing your ap in the trash.

There is a difference not just in different kinds of applicants, but in different kinds of labor markets. The last big recession the old rules were all called off. My husband has often remarked that ten years ago when he was doing the hiring, he got back to every single applicant to tell him or her when a job was filled. But I hear over and over again from very qualified professional people situations in which you never even know if your application has been received, if a job is still open, if it has been filled or set aside.

Our last round my husband and I each independently had the fun experience of being told we unofficially had jobs, then being called a couple days later to be informed someone better fitting just walked in the door. We were decent job candidates, so I think they wanted to hang on to us as prospects, but the pool of applicants was huge so they didn't want to miss out on someone better.

A bit of useful advice I heard years ago is that you don't have the job unless and until you are told when to report. And even then it may not be 100% guaranteed.

I have this crazy theory that those who shirk getting a job due to an unemployment check don't add so much to the economy even after they do get a job.

There several sub-populations of unemployed from which you can expect different sorts of reaction/effects to being in unemployed benefits, or better said: different expected probability of finding the desired job ... You can not compare an unemployed person in their prime (20-35) with one over 50. They need different unemployment benefit policies. Same goes for low vs highly qualified workers....

Exactly. We believe in labor market segmentation, so why don't we believe in unemployment segmentation?

Question, why unemployment weekly payments? Is it more efficient than monthly payments?

Payments I received were biweekly with weekly accounting.

It is more 'efficient' in the sense that you cut off benefits as soon as someone starts working full time (suspended or reduced if work is temporary). First week is a wait week. So benefits turn on slowly and turn off quickly.

How many unfilled job openings are in the U.S. right now? It seems to me we keep missing a pretty big piece of the puzzle, if there are no jobs it really doesn't matter how big the incentive is to look for a position when you're unemployed. You'll get a lot more looking with more incentives, but not more finding.

In a healthy economy, folks who cannot find a position will create one. Maybe we ought to be looking not just at the disincentives to finding work, but the disincentives to creating work -- i.e. obstacles to small business.

Comments for this post are closed