Modern Job-Hunting: a Not so Perfect Game of ‘Perfect Information’
A recently held Philadelphia job fair for ex-offenders got canceled when its doors opened and it became clear that the 3,000-strong crowd of job hunters was about to overwhelm the venue and plunge the fair into chaos.
Only then did it become obvious that the turnout estimate based on the previous year’s attendance figure—1,000 job chasers—was wildly off.
A key contributing cause of the fiasco? As Homer Simpson might put it, ”Too much information!”
More precisely, it was a problem of too many people with too much information—a situation nonetheless idealized and celebrated in classical theories of pure capitalism as the “perfect information” condition.
This means that all the “players” (in this case, the job hunters, the fair sponsors and the participating employers) have the same information about the market, the other players, the “rules of engagement”, etc., and that the information is complete.
This is pretty much as it is for chess players, in virtue of their all having the same knowledge of the permitted moves and positions of all the pieces, which are equally and completely visible to them on the chess board.
Knowing your opponent’s every move in advance would amount to having more perfect information, but then it would make chess more of an exercise than a game, even assuming you made no stupid moves yourself. (Note: The theory of perfect competition assumes all the players are “rational”, which presumably means not stupid.)
In connection with any game of perfect information, you may be forgiven for asking this awkward question: Isn’t it possible for the players to know too much, to have too much information—even though, or because, they all have it?
Imagine a version of “straight poker” (five cards, no draw) in which each player knows every card every other player holds. Wouldn’t that be too much information, at least to the extent of making the game pointless, if not impossible?
Blame What We Know
That perfect information was the root of the attendance spike is clear from what Everett Gillison, deputy mayor for public safety had to say about it: He explained that 25 employers were on hand for this year’s planned fair, for which, in the previous year 1,000 people showed up, with 435 people getting jobs.
He added that, after that success, this time, word of the fair quickly spread on social media and snowballed—headlong, toward an ever-closer approximation of a state of perfect information.
In the pre-Internet, pre-cell phone era, awareness of the job fair would have diffused and rippled much more slowly and imperfectly through the job pool, e.g., by word of mouth, notices read on bulletin boards or newspapers.
In those days, the reality of imperfect information pretty much guaranteed each job hunter that not everyone with similar job qualifications and ambitions would know about any given opening.
These days, to post a job or to hold a job fair amounts to triggering a flash mob, as the news spreads faster than wildfire—indeed, at the speed of light and electrons.
The described job-fair chaos suggests that perfect theories of perfect capitalism’s idealized notion of perfect competition are far from perfect.
In the real world, as the failed job fair evidences, not only is a little knowledge a dangerous thing; a lot of knowledge (in terms of both the quantity of information and the quantity of minds possessing it) can be just as, if not more, dangerous—assuming chaos of the kind that ensued at the fair is not a “good thing”.
The Flash-Mob Model of Job Hunting
In classical perfect-information scenarios, the egalitarian, instantaneous diffusion of market information is presumed to ensure efficient, effective, open and fair competition, e.g., no skewing of economic advantage because of privileged insider information and no misallocation of talent because of ignorance of opportunities.
Well, the flash-mob model of 21st-century job hunting may be fair, but it is hardly efficient and effective. (Note: In case you’ve never heard of or seen one, a flash mob is created through coordinated communication e.g., by means of social media or cell phones, among a large number of people, for the purpose of congregating at a specific time and place for a specific purpose, not necessarily, although frequently, frivolous or nefarious.)
The argument that the flash-mob model’s defects are self-correcting because everyone in the job mob instantly and equally is informed of the cancellation and planned rescheduling of the job fair is not compelling. Using perfect information to correct a perfect-information-induced catastrophe seems terribly wasteful—a failing that is and should be anathema to idealized capitalism.
This failure of perfect-information dynamics makes a case for the good old days, when the ignorance of other job seekers in a tight job market safeguarded orderly hiring, to the degree that at least somebody got a job offer, instead of getting jammed up and turned back at the swarm-packed door.
This modern phenomenon and deficiency of perfect information-driven recruitment has its parallel in modern law enforcement: In bygone eras when not only the “good guys”—the police, but also the “bad guys” had only imperfect information, a convenience store manager could depend on having only one or maybe two shoplifters at a time, filching a couple of items.
Now, in 2013, that shop owner has to brace himself for a coordinated flash mob of fifty cleaning the store’s shelves like locusts on a field trip, having a field day, in somebody else’s field.
Clearly, the old ways and days of imperfect information made things more manageable, more orderly.
To keep up, modern law enforcement agencies are striving harder to make their “game” more like a game of perfect information, e.g., ubiquitous CCTV cameras, monitoring of social media, communications intercepts—perfect, at least for themselves.
Apparently, although the other “players”, the perps in flash mobs, also have comparably perfect information, e.g., know about the cameras and the rest, they are not only fearless, but are also brazen or dumb enough to post their own videos of their crimes on YouTube or Facebook. (So much for the “rational player” models and theories.)
Such is the tangled techno-psychological relationship between perfect information and those who, despite in fact knowing better, try to pull off the perfect crime.
Making Imperfect ‘Perfect Information’ More Perfect: an Oscillating Model
In defense of perfect information, it could be claimed that the problem at the failed job fair was not too much information, but, instead, too little.
If the job hunting ex-offenders knew, in real time, exactly how many people were planning to show up at every moment before the fair, how relatively few employers would be there and how cramped the space would be given the projected crowd size, they would have the information needed to make a more informed decision about whether to bother to attend.
Now, this could get complicated, as some job seekers, becoming discouraged, drop out, to be replaced by others who learning about the dropouts, then decide to attend, thereby swelling the ranks of the job seekers once again, some among whom then become discouraged, drop out, cueing others to attend, etc.
What I’ve just described is a dynamically oscillating system with negative feedback—a system that tends toward an equilibrium value by oscillating around an average projected attendance. That’s about the best that can be expected under such “more perfect information” conditions.
In the basic perfect-information model for a tight job market, it would be 100% certain that 100% of the job hunters will show up in a huge flash mob scrambling for whatever jobs are available—in toe-to-toe fierce, staggering and chaotic competition, if their huge numbers don’t cause a shutdown of the process.
In the second, the oscillating model, although things wouldn’t be so hectic and frantic, they’d be very stressful throughout, e.g., because of the unceasing vigilance, monitoring, instantaneous adapting and re-planning required.
However, neither of these perfect competition models can offer the comforting assurance the “old days” provided.
That was when knowing that finding out about a job didn’t mean everybody else did too. Back then, the imperfect information that employers as well as job hunters had improved your odds: For not only did far fewer job hunters know about the employers and the jobs, but also, employers knew of far fewer job hunters.
In many cases, perhaps the biggest challenge then was just to find out about the job.
Now, the comparable challenge is to find one before everybody else finds it…
….or at least to find the end of the queue, if there’s an end to be found.