The time needed to fill jobs

You can see a tightening labor market, and some of this you might interpret in terms of a growing skills gap:


The national time-to-fill average rose in February to the highest level in 15 years.

Sponsored by the career sites publisher Dice Holdings, the Dice-DFH Vacancy Duration Measure says it took an average of 26.8 working days to fill jobs in February. In January, according to the report, the average (as revised) was 25.7 working days.

“We are continuing to see signs of a tightening labor market,” said Michael Durney, president and CEO of Dice Holdings. “Unemployment rates are declining across several core industries, such as tech and healthcare, and the time-to-fill-open positions has hit an all-time high in the 15 years the data has been tracked.”

Jobs in the financial services sector took the longest to fill, averaging 43.1 days. Healthcare jobs averaged 42.6 days to fill. Both are historic highs for their respective industry sectors.

The full article, by John Zappe, is here, via David Wessel.

By the way, the skills mismatch hypothesis (not a substitute for AD hypotheses, let me stress) is stronger than you think.  Catherine Rampell writes:

Companies themselves are indeed reporting difficulty acquiring talent. Nearly half of small and medium-size businesses with recent vacancies say they could find few or no “qualified applicants,” according to the latest monthly survey of the National Federation of Independent Business. Over the past couple of years, firms have also become increasingly likely to say that “quality of labor” is the “single most important problem” facing their businesses.

I have generally been skeptical of the “skills mismatch” story, mostly because wage growth has been so pitiful during this recovery. If qualified applicants were really so scarce, businesses should be bidding up the salaries of the few hot commodities available. But Steven J. Davis , an economics professor at the University of Chicago’s Booth School of Business and one of the creators of the vacancy duration metric, says it’s unclear how a widespread skills mismatch, if one exists, would actually affect wages. Maybe a shortage of some highly desirable skill would lead employers to bid up the price of that qualification, he says, but that same scarcity might also lead firms to instead settle “for workers who have less of [or] lower-quality versions of the desired skills.” That could, in theory, cause hourly wages to fall overall. Robust research on the connection between skills mismatch and wages is in short supply.

This point from Davis is appreciated only rarely.


The point from Davis sounds a bit unscientific. Let's see some numbers.

I think his point makes sense.

Your argument that Davis' postulate that Davis says needs data to test is not backed by data, so your claim about Davis point is unscientific.

Perhaps I'm dense, but I don't understand Davis' point. Have you seen (per the eponymous Second Law) any attempt at formally modeling such behavior?


Also, an alternative explanation for the above time series: Davis' other work on falling 'recruiting intensity'. Has been discussed previously on MR:

Key figure:

I'm also not sure I understand it. I think it's like this: (1) The company really wants someone with 20 years experience leading the development of mobile apps, but (2) That person doesn't actually exist. So instead they hire someone with 20 years experience developing desktop apps. Or someone with 7 years experience leading the development of mobile apps. And either way, they pay them less than they would have paid a hypothetical ideal candidate.

This aligns pretty well with the common observation that job postings have become hyper-specific over the past few years.

Dear small and medium-size businesses with recent vacancies,

That's why on the job training was a thing once.upon a time.

How many people are you training, if you don't mind me asking?

7, all in India.

Over the past couple of years, firms have also become increasingly likely to say that “quality of labor” is the “single most important problem” facing their businesses....But Steven J. Davis , an economics professor at the University of Chicago’s Booth School of Business and one of the creators of the vacancy duration metric, says it’s unclear how a widespread skills mismatch, if one exists, would actually affect wages. Maybe a shortage of some highly desirable skill would lead employers to bid up the price of that qualification, he says, but that same scarcity might also lead firms to instead settle “for workers who have less of [or] lower-quality versions of the desired skills.”

Look, the fact that firms say they want "skills" and "talent" in some stupid survey doesn't mean jack shit unless they're explicitly defined and translated into bidding wars that increase wages for those particular "skills" or "talent". The fact that this is not the case and that firms are instead hiring workers who don't have these "skills" suggests that "skills" and "talent" are meaningless buzz words and idioms for these firms, like high school football coaches demanding "110%" from their players on Friday night.

So where would the money come from to bid up wages? I have said many times that the US and most western democracies have priced themselves out of the market with rare exceptions. If costs increase they won't have the need to hire.

Wages are heavily built by price levels. The fact is, they are already paying a pretty good wage.

One of the constant challenges that I face is not finding work to do; there is a very very high demand for my services and expertise. It is finding people willing and able to pay for the work that I do. We had this discussion yesterday at work. One of the guys is bringing in all these leads for work. Great. So do they have money, do they know what it costs to have us spend half a day on something? Usually not.

There are a few trends here.

1. Demographics. A very large number of qualified and experienced people are retiring.

2. Education costs and debt. A new hire comes with high costs before they even know how unqualified they are.

3.Two tier economy. Some sectors are paying very well setting the price level for certain skills very high. They can afford them because they are very profitable. Outside of those sectors those wage levels are simply put impossible.

4. Foreign competition. Everyone competes with a billion Chinese workers and a billion Indian workers willing to work twice as hard for half the wages. In most sectors of the economy that is the wage level they can pay.

5. For a whole list of reasons training is not done by businesses any more. From poaching, low margins, short term project based employment, etc.

6. Every employee trained or not comes instantly from the first day with substantial costs. Health care, regulatory requirements. Someone has to be damned good to pay their way right off the start.

None of these issues are a problem if there is 4% growth. At 2% growth they become a choke point.

Re: Every employee trained or not comes instantly from the first day with substantial costs. Health care, regulatory requirements. Someone has to be damned good to pay their way right off the start.

Nothing new under the sun there. For skilled work especially benefit costs have been baked in the cake since well before most of us were born. But it was never the case that very many people could "pay their own way": a three month to six month (maybe even one year) training gap has always been existed, depending on the complexity and skill level of the work.

What is the time required from starting schooling to bring able to take on any project or task in your field? In mine by the third year you are marginally useful, and 8-10 years can take on most anything. I talked to a mechanical engineer a while ago, I asked him that question and he said 10 years.

I see the employment situation as described. There is the normal turnover in staffing, but there are lots of experienced, really core people retiring as the baby boom generation gets older. Shockingly few have trained their replacement. Last year I asked everyone i ran into who was nearing retirement whether they have trained their replacements, and almost unanimously the answer was no. Often these were jobs that are very specialized, the field is miniscule, but without things grind quickly to a halt.

My trade has a tradition of training, but seemingly no longer. Hard to find people who are willing to take responsibility, willing to work hard, even willing to put in the hard work of learning.

Raise wages? From what? Already we are facing the situation where an increase in our costs ended up meaning things just are not done. My clients have come off of a string of brutal years, barely surviving as it is, and simply won't pay more.

In most cases the skills are available elsewhere for a lower price. Education debt will end up pricing quite a few qualified people out of the market.

The long lead times to hire simply means that things are not done.

Hard to find people who are willing to take responsibility, willing to work hard, even willing to put in the hard work of learning.

My experience in the work place over the last 30 years has been that many employers do not encourage much if any mentoring.

Also seems that very few experienced and competent people are willing to take on the responsibility for and the hard work of mentoring. There is little career upside to being a good mentor, and potentially a LOT of downside. And if you do mentor and are good at it, as your reputation for that spreads, you will find many eager younger people who want to talk with you, taking even more time.

That may be why, in part, the executive and business coaching business has grown.

Derek, what kind of trades are these? Who is the target pipeline? Are we talking baby-faces fresh out of school, mid-career tradesmen? What? I can tell you I work in a corporate Fortune 50 and the guy in the cube next to me pulls 60-70 hour weeks, every week. His buddy runs a home electrical company, and needs workers since his are retiring. He offered to pay my Fortune 50 coworker a stipend and all of his tuition expenses for 2 years of schooling, and starting him out at $75k a year at the end of those 2 years (factoring in overtime for his expected 50-60 hour week). That's what a "Skill Shortage" looks like to me.

In contrast, skills shortage to our Fortune 50 means....our department lost two people. We are slated to hire one additional person. We interview no one for six months and make everyone else do the work. We complain that no one is "taking ownership" of our laundry list of issues. We are told the only approved training is the kind that costs $0, and the Account Managers tell us training is our responsibility, in our own time, at our own expense. Compare/Contrast with the above.

What I do is the equivalent to a service electrician. Most electrical work is installation, which has it's demands but requires less experience to be competent. It sounds like two years is what we would call a pre-apprenticeship where all the trade training is done in one shot as opposed to sessions over the apprenticeship period. We have an apprenticeship program, electrical is 5 years worth of time and about 40 weeks of training. What I do, refrigeration service, is 5 years with 28 weeks apprenticeship training, at the end both can do regulated work.

A second year apprentice in my trade costs their employer money. A second year electrical apprentice is a valuable asset for an installation crew.

It's the "purple squirrel" phenomenon and it's getting more severe all the time. There is really no meaning to "skills gap" when the employers are going crazy over zero training requirements, zero risk, cultural fit, retention, docility, age.

Economists need to revise their terminology and stop talking about a "skills gap". A skills gap can be rectified by traditional training.

When an employer wants a "digital native" (ie. young, cheap, lots of overtime, no family) with 2 years of experience in building SaaS applications for the insurance industry using the latest and greatest webapp framework it's not correct to talk about a "skills gap".

That has to be the stupidest theory for how a rational actor will respond to a supply shortage that I've ever heard.

Fifteen years seems like a long time, but this data series really only shows two economic cycles. Vacancy duration falls during recessions and increases during recoveries. Last cycle, duration rose from 2004 to 2007 before falling from 2007-2009 due to the recession. In the alternative universe where there was no Great Recession, would duration have flattened or continued to rise? There was some flattening in 2013-2014 before durations resumed increasing. How do we know that this is not the normal pattern as long as the economy avoids recession? What were durations like before the 2001-2002 recession, when the graph starts?

Could also be about the rise of professional temp work. Makes it less urgent to fill any given position.

it’s unclear how a widespread skills mismatch, if one exists, would actually affect wages

Indeed! CEO wages have been rising monotonically, but there is no evidence that "CEO skills" (whatever they may be) are in short supply, nor that there are few people around who are "CEO material" (whatever that might mean.)

It may just be that employers are asking too much of their hires to ever be satisfied by the average candidate they interview. The MBA-types have so drilled into their heads the notion that a bad hire will destroy their companies that it has made employers shoot for the moon instead of taking a chance with ordinary, competent-looking, candidates.

It's encouraging that Cowen is less prone to mood affiliation; he is a scholar, first and foremost. The skills gap purveyors remind me of the authors of the study that concluded that the spike in employment in 2014 was attributable to the cutoff of unemployment benefits at the end of 2013. I don't doubt that skills are relevant, as are policies that discourage work, but why latch onto politically convenient explanations for complex phenomena. A less politically convenient explanation for the "skills gap" is that employers don't wish to invest in training the way they once did, in part because employment is no longer a lifetime commitment (from either the employer or the employee), in part because of recent experience and a hangover of pessimism about economic stability, in part because, well, there are likely many parts. When I was a young lawyer in the 1970s, the best law firms only recruited at the law schools, eschewing experienced lateral hires (those without a "skill gap") because the firm wanted its lawyers to be trained in the firm way. Sure, there was a certain arrogance involved, but there was good reason for wanting the firm's lawyers to reflect the firm's approaching to lawyering. That seems quaint today, as law firms raid one another to fill the "skills gap" for whatever the latest hot market for lawyering skills (intellectual property, health care, whatever). And so it goes from law to banking to programming and down the line. The relationship between employer and employee has changed (indeed, the growing preference is for a relationship other than employer-employee), some say for the worse, others say for the better. My view is that we should adopt policies that encourage employment, and terminate policies that discourage employment, that we should adopt policies that encourage training, mutual commitment, and long-term employment, and terminate policies that discourage training, commitment, and long-term employment. That's a tall order, much taller than attributing unemployment to the "skills gap".

I think the fact that the cost of listing vacancies has effectively dropped to zero (almost every large company now lists their openings on their own website) means that there is every incentive for managers to open up a set of standard vacancy announcements covering their entire workforce, so they can hire immediately whenever a good candidate comes along. When the chance comes to get someone good, managers want to make an offer that day, not wait around for a week or two for a new vacancy announcement to make it through their company's HR department. This is going to skew this metric pretty badly.

Does this happen in companies with an HR department? There is no power to hire outside the standard HR process at my company, or at other companies with which I'm familiar (this applies to professional positions up to the executive VP level -- not sure what the process is there).

It rings true to me. I am involved in hiring tech talent at a Fortune 50 company. There's requisition trading and borrowing between managers and other ways to jump on a good candidate if need be.

It's not uncommon for job listings to be pit up (with every intention to hire) but then orders for a hiring freeze, or at least a hiring slowdown will come down from on high. I worked at a firm until last year where every year there was a hiring freeze (on external hires) the last two or three months of the year, yet jobs were still being posted.


Metrics are almost universally more volatile than the phenomena they purport to measure.

Ultimately, if there is a high demand for something, and a low supply, then the price for that something will rise, which increases the supply as more people want to profit off the higher price. This is basically the law upon which modern economics is built.

So, either:

(A) business are too dumb/greedy/stubborn to allow the cost of skilled labor to increase, or
(B) There really is no "skills gap", and instead businesses use that term to argue for an increase in foreign (and cheaper) labor, or
(C) The modern economic theory is no longer valid.

(B) seems most likely.

Of course no moral, intelligent person could argue against (B)

Why not? Because it's foreign? or because it's cheap? No one complains about immigrants coming in at market rates; people complain about immigrants coming in for the express intent to depress wages; after all, as a supplier, if I dumped a lot of cheap commodity into an economy in order to depress the price, isn't that a violation? Don't we recognize that as a market distortion?

Supply-siders are all about sharing the wealth...when it's the wealth of the run below them getting shared. When it's about sharing their wealth - things like estate taxes and whatnot - well, that's just tyranny, and communism, and fascism, and immoral.

In my experience, it's a mix of A and B. I work in the high-tech industry in a specialized field, and we've had huge problems hiring. Not because the people weren't available, though; we've interviewed a number of people that would have worked great. Too bad we couldn't get management to approve their salary requests.

Instead, they've done things like contract work out to Wipro. We warned them up-front that they don't have the best track record, and the project would turn out badly. It has. It's going to cost us a lot more to fix than it would have cost to do right the first time.

And as for A, I've actually had managers complain about having to manage engineers that make more than they do.

Isn't this just reflecting job openings?

If more jobs are advertised, I would expect the mean vacancy time to increase in the absence of a corresponding increase in the number of job seekers, growing skills gap or not.

I am certain this has nothing to do with millions of inexperienced 20 - somethings entering the workforce and millions of very experienced boomers leaving it... Demographics seem to be an under - appreciated driver of macro trends overall (see the argument over the possibility of 4% growth in GDP). Demand for skill does drive supply, but adjustment can be costly and slow.

...And, as always, skip over the fewer-millions of Generation X who should be at their peak earning/experience potential.

Maybe it's a modern romance theory. It's not that there's a growing skills gap, it's that there's a perceived growth because online dating, er, job applications, increase the perception there's something better out there.

There is a skills gap in my area. I want to hire someone twice a smart as Einstein and first rate communication skills. He has to hit the ground running as an expert in statistics, software design, and finance.

Unfortunately I could not find such a candidate, so I settled for real people.

If there was a growing skills gap, could this not be measured by observing whether expenditures on employee training are increasing?

What do you think this is, Germany?

Not necessarily, if there's a significant lag between training and reaping the benefits of training. If the problem is that there aren't enough people who were trained five years ago, that's not something I can fix today. And investing in training today might not help me if labor mobility is so high I'm unlikely to be the recipient of the labor of my newly trained employee when the training actually starts to pay off.

Decent multi-year labor contracts might help solve this problem. "We will pay you at least $X over four years and cannot fire you, but in addition to your labor we get an enforceable non-compete."

A lot of training is done informally. It's not always an explicit budget item.

Maybe the "purple squirrel" phenomenon can also explain stagnant wages.

If _all_ employers are looking for a very specific set of qualifications, then the small set of people who match those qualifications will not have many employment options, since those people can only go to the (usually) small number of employers that seek those qualifications.

That's very close to Davis' point. It's like the complement of the set. I think Davis is saying that employers honestly need a purple squirrel, and aren't just kidding themselves. So when they have to settle for a regular brown squirrel, they pay them less.

Perhaps they need a purple squirrel, but are only willing to pay for brown squirrels, though? They prize purpleness, but feel like they shouldn't have to pay for it?

In the past, people paid for squirrels at one squirrel price, but a few turned out to be purple and they were great. So now, employers all want purple squirrels, but they remember the days when they paid one price for squirrels, and so all want to just keep paying the regular squirrel price, but get more purple squirrels somehow?

At the risk of straining the conversation to the breaking point, it's possible that both (1) the firm needs a purple squirrel, and (2) purple squirrels don't exist. Paying more for this opening doesn't make more purple squirrels exist, at least in the short term for an individual employer.

That said, I'm inclined to believe firms don't need purple squirrels nearly as often as they think they do. And that firms are bad at identifying the purple squirrel even though they are very confident in their look-you-in-the-eye processes. It's all "guess the dress code" BS...

That's my perception of the matter.

I'm not going to argue there's no real skills gap, but to a certain extent it does seem like employers are basically screaming "I want a Pony!"

Of course, every employer wants the PERFECT candidate who can show up and start doing everything on day one with no training. Obviously, that's about as hard to find as a unicorn.

However, there is also the underlying information problem caused by the student loan system. The student loan system makes loans available, at the same interest rates, to any student studying any major. This obscures information that would otherwise inform college students about what skills are in demand in the market which casues skills mismatch. If student loan interest rates were set by the market, they would be much lower for majors that were in demand and much higher for majors where there was an oversupply of workers.

The lead time in the system (which amounts to years) is such that by the time students enter the work force those signals would be very stale.

Do the skills in demand in the workforce really change that rapidly? The body of people who have certain skills don't just die off en masse, so the supply will remain fairly steady. Short of something like a major war, I don't think the demand for doctors or engineers is going to fluctuate that rapidly.

Re: Do the skills in demand in the workforce really change that rapidly?

Good grief, yes. Scads of people went into IT in the late 90s-- only to be greeted by the tech bust in the early 00s (followed by the outsourcing of entry level jobs in the years after)

If so, that could also be addressed. It is fairly easy for a diligent person to complete 120 credit hours in two years of linear time, and with the right financial motivations maybe students would focus on getting the degree and getting out rather than socializing. An employer will want to hire several months in advance at least, so I think we are well within product lifecycle planning norms.

Unless socializing includes part of the screening process itself.

I can guess why a good deal of recruitment happens through "old school" networks. (Not to mention many, perhaps a majority of, people met their romantic partners in high school or college.)

There are a few things that conspire against those who train.

1. It costs money.

2. The started trainees to trained is I would say 4:1. Lots of people want to be trained, but in anything of any complexity few are up to it or willing to put forth the effort to be trained.

3. Well trained employees are ripe for the plucking. It is far cheaper to poach, pay slightly higher wages, than to train. A company who trains very well will train lots of people for someone else.

4. Most companies who train hire trainees when they have a project or similar. When times are slow trainees aren't hired. It has been slow for almost a decade. That is probably two generations of new trainees that haven't happened.

"Mean Vacancy Duration"... What's the median? I imagine there are a few job postings that stay up for years, such as "We need someone with extremely advanced IT knowledge and 15 years experience to work for 35k/yr."

Those are simply a way to say we are hiring. But we will hire who we want and since no one fills the qualifications no one can complain.

The employers are more or less telling the truth. We have an open position right now and very much want to fill it, I've personally interviewed more than a dozen candidates in person or on the phone over a several month period. Most of them have a complete lack of talent - if they had any talent whatsoever training would not be an issue. Our demands are probably not unreasonable - all the existing staff are pretty good, I've interviewed students at a decent state school who knew most of the answers well enough, we've had a few candidates reject our offer for a slightly different offer elsewhere, and we even filled another similar position. Wages are OK on an absolute scale, certainly above 100k which is not bad in our area.

Management doesn't want to get in a bidding war as they would have to pay all the existing staff more, we would probably lose the war to companies that just print stock in unlimited quantities, the position isn't worth that much compared to just leaving it open, and some people just want to do work at slightly more popular companies even at a lower standard of living. So basically the market will clear, but it takes two years to clear as we need to recruit an intern during intern season or wait for someone's wife to move to the area or something that motivates someone actually qualified to apply.

I hear a lot about the rise of non-competes in all manner of industries. Surely this could be a contributing factor? If I want to hire someone with X skills away from their current employer, even if I'm willing to pay them a bundle more than they currently earn, with non-competes in place you would see a situation where the wages weren't showing a rise and companies in my position would complain they can't fill positions.

Maybe firms are offering too little money, and so qualified applicants decide to spend their days in easier jobs or working part time.

It used to be that firms would hire a broadly educated individual and train them. Now, firms wish that they can hire someone who will hit the ground running on day/month one. That's not a skills gap, it's a gap in realism, and it's having horrible effects on the ability of firms to build the workforce of the future.

I am never concerned about jobs or anything like that because of OctaFX broker and their brilliant service, it’s especially due to their low spread of 0.2 pips, high leverage up to 1:500 and solid cTrader platform where all my trades are executed perfectly, hence one can’t really worry about working when there is so much support by any company, it just makes me believe that I belong here and that’s why I am successful thanks to their service and facilities.

Check out Peter Cappelli's Why Good People Can't Get Jobs. Among other things, he points out that one way or another employers will pay for training. If it's the prospective employees who have to shell out for training, (1) they'll demand it back in their wages and (2) employers will sometimes not be able to find enough skilled people.

Comments for this post are closed