Congestion and signaling in the job market, as the ratio of applications to positions continues to rise

 Aki Ito, at Business Insider, writes about how the number of job applications per position is growing, and how there's some exploratory use of signaling of interest through job sites that allow a small number of such signals.

How tech broke the job market
Applying to a job in 2025 is the statistical equivalent of hurling your resume into a black hole. By Aki Ito

"To see how bad it's gotten, I asked Greenhouse, one of the leading providers of hiring software, to take a look at their data. Last quarter, the average job opening received 242 applications — nearly triple the amount in 2017, when the unemployment rate was at a comparable level.

 

"Nobody's happy with the current situation," says Greenhouse CEO Daniel Chait. "Something broke in the technology." 

"This isn't the first time a market's grown so overcrowded it stopped functioning. Economists even have a name for it: congestion. Big markets hold the promise of creating better matches, but they also tend to devolve into total chaos.

"Congestion is the bane of a lot of markets," says Alvin Roth, a Nobel Prize-winning economist at Stanford who's helped design programs to better match students with schools, organ donors with patients, and hospitals with new doctors. "Successful marketplaces have to fight hard to defeat congestion."

...

"The forces that make it cheap to send more applications are working faster than the forces that allow you to quickly process many applications," says Roth. "We're deep into congestion."

...

[There is] "a new website where candidates can manage their applications to Greenhouse's clients. There, it introduced a feature called Dream Job, which lets people mark one application a month as a job they especially want. The idea is that recruiters don't just want qualified applicants. They want to know — amid the sea of people applying with a single click — who's actually serious enough that they'd likely accept an offer.

"Online daters might recognize the concept as the "rose" on Hinge or the "super like" on Tinder — gestures borrowed from a landmark study in market design. Dream Job launched in June, and the early data is promising: Employers have been five times more likely to hire Dream Job applicants than standard ones.

"Other intermediaries of the job market are trying their own fixes. LinkedIn, for instance, introduced its own "rose," called Top Choice, to its premium members (Top Choice candidates, the platform says, are 43% more likely to get a recruiter message). It also shows people whether they're a high, medium, or low match for the roles they view ("try exploring other jobs," it gently advises low-match candidates). And this year it's been testing daily limits on Easy Apply submissions."
 

Congestion in the job market, AI version

 The Atlantic has this story on the job market, that contains a nice line...

The Job Market Is Hell.  Young people are using ChatGPT to write their applications; HR is using AI to read them; no one is getting hired.   By Annie Lowrey

“ What Bumble and Hinge did to the dating market, contemporary human-resources practices have done to the job market. People are swiping like crazy and getting nothing back.”

Congestion in online labor markets: too many applications

 As online job ads make it easier to submit chatbot-assisted applications, companies are becoming overwhelmed.

The WSJ has this story:

How to Navigate the Jungle of Online Job Postings
Companies are rethinking online job applications, seeking quality over quantity  By Callum Borchers

"You probably haven’t looked for a job in a newspaper’s classified pages since the Bush administration—possibly the first one. It could be worth reviving this old-school strategy because many of the listings offer a way to bypass those dreaded online application portals.

...

"Companies fed up with the low-quality, sometimes fraudulent submissions that flood applicant-tracking systems are reaching back in time for hard-to-hack recruiting methods. Classified ads are just one tack.

"Others include: leaning harder on references; making application forms so cumbersome that only serious candidates will complete them; and posting openings on niche job boards instead of the most popular ones."

Deceased organ allocation: deciding early when to move fast

The deceased donor waiting list for kidneys to transplant is congested: offers, which take time to evaluate, are often rejected, while cold ischemia time accumulates.

 Here's a paper just published in Transplantation, in which we suggest new ways to detect organs that will be hard to match, and which might therefore be expedited through the allocation process (to get more quickly to patients who will accept them).

Insights From Refusal Patterns for Deceased Donor Kidney Offers, by Guan, Grace MS1; Neelam, Sanjit MS2; Studnia, Joachim MS2; Cheng, Xingxing S. MD, MS3; Melcher, Marc L. MD, PhD4; Rees, Michael A. MD, PhD5,6; Roth, Alvin E. PhD7; Somaini, Paulo PhD8; Ashlagi, Itai PhD1
Author Information
Transplantation ():10.1097/TP.0000000000005434, May 21, 2025 

"Background.
The likelihood that a deceased donor kidney will be used evolves during the allocation process. Transplant centers can either decline an organ offer for a single patient or for multiple patients at the same time. We hypothesize that refusals for a single patient indicate issues with individual patients, whereas simultaneous refusals for multiple patients indicate issues with organ quality.

Methods.
We investigate offer refusal patterns between January 1, 2022, and December 31, 2023, using Organ Procurement and Transplantation Network data. We aggregate refusals at the same timestamp by a center and define a multiple patient refusal as >1 or >5 patients simultaneously refused. We report the refusal codes associated with single and multiple patient refusals and the nonutilization rate after receiving single and multiple patient refusals by cross-clamp.

Results.
Patient-related refusal reasons are more commonly single patient refusals, whereas organ-related refusal reasons are more commonly multiple patient refusals. Multiple patient refusals before cross-clamp are associated with nonutilization, but single patient refusals are positively correlated with utilization. The nonutilization rate was 28% for organs without pre-clamp refusals, 35% with a single center sending a multiple patient refusal, but only 12% with a single center sending a single patient refusal.

Conclusions.
The risk of nonutilization can be assessed early in the offering process based on the number of single and multiple patient refusals received by a specific time (e.g., cross-clamp). Understanding refusal patterns can guide the development of transparent protocols for accelerated placement."

 

Dating sites as matching markets: Bumble reimagined

The NYT interviews Whitney Wolfe Herd, who co-founded Twitter in 2012, started Bumble in 2014, stepped down as Bumble CEO and is now resuming that position, amidst some general malaise among dating apps, reflected in stock prices and drop-off in younger participants.  The interview is wide ranging and interesting. I'll excerpt two market design observations, both concerned with congestion--i.e. with the difficulty of curating and finding matches in a large market.

Here's the NYT interview:
Can Whitney Wolfe Herd Make Us Love Dating Apps Again?  By Lulu Garcia-Navarro

"The next era of Bumble, you had a lot of growth during the pandemic when everyone was stuck on their apps. It was a huge moment. You go public in 2021, ring the bell, baby on your hip, and the very next year user growth starts to slow down. What do you think was happening? My opinion is that I ran this company for the first several years as a quality over quantity approach. A telephone provider came to us early on. They said, “We love your brand, we want to put your app preprogrammed on all of our phones and when people buy our phones, your app will be on the home screen, and you’re going to get millions of free downloads.” I said, “Thank you so much but no thank you.” Nobody could understand what in the world I was doing, and I said it’s the wrong way to grow. This is not a social network, this is a double-sided marketplace. One person gets on and they have to see someone that is relevant to them. If you flood the system just endlessly — you’re not going to walk down the streets of New York City and want to meet every single person you pass. Why would you assume that someone would want to do that on an app? This is not a content platform where you can just scroll and scroll and scroll and scale drives results. What happened was, in the pandemic and throughout other chapters, growth was king. It was hailed as the end all be all.

...

"You’re quite bullish on A.I. I’ve heard you talk about it. How are you imagining A.I. functioning in this next iteration of the app? Let’s say we could train A.I. on thousands of what we perceive as great profiles, and the A.I. can get so sophisticated at understanding: “Wow, this person has a thoughtful bio. This person has photos that are not blurry. They’re not all group photos. They’re not wearing sunglasses. We can see who they are clearly and we understand that they took time.” The A.I. can now select the best people and start showing the best people the best people and start getting you to a match quicker, more efficiently, more thoughtfully. The goal for Bumble over the next few years is to become the world’s smartest matchmaker. This is beyond love. We have a friend product with a very broad member base, and it’s really beautiful."

Freezing and thawing organs for transplant moves one step closer

 In March, surgeons at Mass General Hospital thawed and transplanted a frozen pig organ into a pig.  The challenge of freezing and then thawing an organ back to life, so that it can be stored until an appropriate transplant can be arranged, is one of long standing. The difficulty is that during both freezing and thawing, there is a danger of ice crystals forming inside the cells, which would destroy them.

Here's a NYT article that explains why being able to freeze and then successfully thaw organs could help relieve the congestion in kidney transplants for humans.

This Kidney Was Frozen for 10 Days. Could Surgeons Transplant It?
Scientists developed a way to freeze a large mammal’s kidney, which could ease organ shortages in the future. First, they had to see if their method would work in a pig.   By Gina Kolata

"the promise from freezing and storing organs is great.

"There is a severe and ongoing shortage of kidneys for transplants — more than 92,000 people are on waiting lists. One reason is that the window of 24 to 36 hours is so brief that it limits the number of recipients who are good matches.

"How much better it might be to have a bank of stored, frozen organs so an organ transplant could be almost like an elective surgery.

"That, at least, has been the decades-long dream of transplant surgeons.

But the attempts of medical researchers to freeze organs were thwarted at every turn. In many cases, ice crystals formed and destroyed the organs. "

 

HT: Colin Rowat

#############

Here's an earlier post, about a 2017 paper that turns out to have set some of the goal posts:

Monday, June 12, 2017  Organ preservation could bring big changes to transplantation

Transplantation would be a lot less hectic if organs could be preserved. Here's a 42-author paper (the biggest coauthorship I've been involved in) that discusses some of the possibilities.

The promise of organ and tissue preservation to transform medicine 
 Sebastian Giwa, Jedediah K Lewis, Luis Alvarez, Robert Langer, Alvin E Roth, George M Church, James F Markmann, David H Sachs, Anil Chandraker, Jason A Wertheim, Martine Rothblatt, Edward S Boyden, Elling Eidbo, W P Andrew Lee, Bohdan Pomahac, Gerald Brandacher, David M Weinstock, Gloria Elliott, David Nelson, Jason P Acker, Korkut Uygun, Boris Schmalz, Brad P Weegman, Alessandro Tocchio, Greg M Fahy, Kenneth B Storey, Boris Rubinsky, John Bischof, Janet A W Elliott, Teresa K Woodruff, G John Morris, Utkan Demirci, Kelvin G M Brockbank, Erik J Woods, Robert N Ben, John G Baust, Dayong Gao, Barry Fuller, Yoed Rabin, David C Kravitz, Michael J Taylor & Mehmet Toner

Nature Biotechnology 35, 530–542 (2017) doi:10.1038/nbt.3889
Published online 07 June 2017

##

Here's the Google Scholar link, which also includes links to the subsequent literature:

The promise of organ and tissue preservation to transform medicine

S Giwa, JK Lewis, L Alvarez, R Langer, AE Roth… - Nature …, 2017 - nature.com

Cited by 552 Related articles All 17 versions Web of Science: 371

 

Mike Ostrovsky on congestion pricing (podcast)

Congestion pricing: as it's happening:

Congestion Pricing: Economics, Theory, Reality

March 29, 2025 • 57 mins
with @mostrovs @skominers @rhhackett

Welcome to web3 with a16z. I’m your host Robert Hackett, and today we’re talking about congestion pricing — an area of mechanism design that’s aimed at alleviating something everyone hates: traffic.

Now you may have heard this term recently since New York adopted its own version of congestion pricing at the beginning of the year. This is the first program of its kind in the U.S. — and it’s got supporters and detractors. We’ll talk about that, and we’re also going to talk about much more.

In the first part of today’s episode we’ll trace the history of the economic ideas that got us here. In the middle, we’ll dig deeper into the details of putting congestion pricing into practice, plus technological alternatives. And in the final part, we’ll explore parallels to — and implications for — crypto networks.

Our guests are Michael Ostrovsky, a Stanford Economics Professor who specializes in this area and who has done research on congestion pricing in New York. We’re also joined by a16z crypto Research Partner Scott Kominers, who is a Professor of Business Administration at Harvard Business School where he teaches market design and entrepreneurship.

Resources:

• Michael Ostrovsky's paper on congestion pricing in New York City (from before the launch, foreseeing its issues): https://web.stanford.edu/~ost/papers/nyc.pdf
• Michael Ostrovsky's thread that went viral on X shortly after the debut of congestion pricing in New York, discussing the post-launch evidence, his team's data collection efforts, and the link between observed data and predictions in the above paper: https://x.com/mostrovs/status/1876798157595476420
• Two of Ostrovsky's earlier theoretical papers on the topic: (1) https://web.stanford.edu/~ost/papers/complementarity.pdf, (2) https://web.stanford.edu/~ost/papers/sdc.pdf
• Economist William Vickrey's influential paper on congestion pricing: Vickrey, W. S. (1969). Congestion theory and transport investment. American Economic Review 59 (2), 251–260. https://matthewturner.org/ec2410/readings/Vickrey_AER_1969.pdf

Donald Shoup (1938-2025) led the war on (too much) free parking

 Here's his WSJ obit:

Donald Shoup, a Parking Guru Who Reshaped the Urban Landscape, Dies at 86
An economist at UCLA, Shoup said free parking carries a high cost, which is borne by everybody, By Jon Mooallem 

“Nothing is more pedestrian than parking,” he often joked. Everyone else is focused on traffic, Shoup told the website Streetsblog. “I thought I could find something useful if I studied what cars do for 95 percent of the time, which is park.”

...

"In the mid-1960s, he was working in Midtown Manhattan while completing a Ph.D. at Yale and suddenly noticed a paradox he couldn’t make sense of as an economist, but which everyone took for granted as human beings. Up and down West 44th Street, “almost all cars were parking for free on some of the most valuable land on earth,” Shoup recalled on the “Curb Enthusiasm” podcast. 

...

“We have expensive housing for people, and free parking for cars! We have our priorities the wrong way around,” Shoup cried out...

 ...

“A surprising amount of traffic isn’t caused by people who are on their way somewhere,” Shoup concluded in a 2007 New York Times opinion column. “Rather, it is caused by those who have already arrived.” 

...

"Shoup’s policy prescriptions were straightforward: Get rid of minimum parking requirements; bring the price of on-street parking in line with demand, enough to maintain one or two empty spots on every block; and funnel the resulting revenue into upkeep and other public services for the immediate area, creating what Shoup called a “parking benefit district,” to bring residents and local businesses on board.

"There were already some successful test cases of these reforms when “The High Cost of Free Parking” was published, most notably in Pasadena, Calif., where reinstalling parking meters was key to revitalizing a historic shopping district. And Shoup considered the proposals in his book so sensible and self-evident that, he later recalled, he naively assumed his vision would immediately become a reality. “I thought the world would change next month,” he told the podcast, “The War on Cars.”

#######

HT: Atila Abduldakiroglu 

Stable Matching with Interviews, by Ashlagi, Chen, Roghani and Saberi

 Job applicants can now easily submit many job applications, and so interviewing applicants, which is time consuming, has become a major source of congestion in many labor markets.  But even in labor markets that use a clearinghouse to process offers and acceptances (like the market for medical residents in the U.S., the NRMP match) interviews are often organized in a decentralized manner. Here's a paper that tackles the question of how to organize an interview match, under some assumptions about what kind of information is obtained in interviews.  Two approaches are considered: an 'adaptive' algorithm that takes into account the results of previous interviews in assigning subsequent interviews, and a 'non-adaptive' algorithm that matches candidates to interviews before any interview results are known.

Stable Matching with Interviews, by Itai Ashlagi, Jiale Chen, Mohammad Roghani, and Amin Saberi (all at Stanford)

Abstract
"In several two-sided markets, including labor and dating, agents typically have limited information about their preferences prior to mutual interactions. This issue can result in matching frictions, as arising in the labor market for medical residencies, where high application rates are followed by a large number of interviews. Yet, the extensive literature on two-sided matching primarily focuses on models where agents know their preferences, leaving the interactions necessary for preference discovery largely overlooked. This paper studies this problem using an algorithmic approach, extending Gale-Shapley’s deferred acceptance to this context. Two algorithms are proposed. The first is an adaptive algorithm that expands upon GaleShapley’s deferred acceptance by incorporating interviews between applicants and positions. Similar to deferred acceptance, one side sequentially proposes to the other. However, the order of proposals is carefully chosen to ensure an interim stable matching is found. Furthermore, with high probability, the number of interviews conducted by each applicant or position is limited to O(log^2 n).
"In many seasonal markets, interactions occur more simultaneously, consisting of an initial interview phase followed by a clearing stage. We present a non-adaptive algorithm for generating a single stage set of in tiered random markets. The algorithm finds an interim stable matching in such markets while assigning no more than O(log^3 n) interviews to each applicant or position. "

Signaling to decongest job applications and interviews: update on the market for medical residents

 Signaling for interviews is evolving in the market for new doctors, i.e. for medical residencies.  Some specialties are allowing a relatively small number of signals (as in Economics), while others are moving towards many signals, which are functioning as soft caps on the number of applications, since many residency programs in those specialties won't give an interview to someone who doesn't signal them.

Ozair, A., Hanson, J. T., Detchou, D. K., Blackwell, M. P., Jenkins, A., Tissot, M. I., Barrie, U., & McDermott, M.  W. (2024). Program Signaling and Geographic Preferences in the United States Residency Match for Neurosurgery. Cureus, 16(9), e69780. https://doi.org/10.7759/cureus.69780