Monday is easily the slowest night of the week for these prost.i.tutes. Fridays are the busiest, but on Sat.u.r.day night a prost.i.tute will typically earn about 20 percent more than on Friday.

Why isn't the busiest night also the most profitable? Because the single greatest determinant of a prost.i.tute's price is the specific trick she is hired to perform. And for whatever reason, Sat.u.r.day customers purchase more expensive services. Consider the four different s.e.xual acts these prost.i.tutes routinely performed, each with its own price tag:

It's interesting to note that the price of oral s.e.x has plummeted over time relative to "regular" s.e.xual intercourse. In the days of the Everleigh Club, men paid double or triple for oral s.e.x; now it costs less than half the price of intercourse. Why?

True, oral s.e.x imposes a lower cost on the prost.i.tute because it eliminates the possibility of pregnancy and lessens the risk of s.e.xually transmitted disease. (It also offers what one public-health scholar calls "ease of exit," whereby a prost.i.tute can hurriedly escape the police or a threatening customer.) But oral s.e.x always had those benefits. What accounted for the price difference in the old days?

The best answer is that oral s.e.x carried a sort of taboo tax. At the time, it was considered a form of perversion, especially by religious-minded folks, since it satisfied the l.u.s.t requirements of s.e.x without fulfilling the reproductive requirements. The Everleigh Club was of course happy to profit from this taboo. Indeed, the club's physician avidly endorsed oral s.e.x because it meant higher profits for the establishment and less wear and tear on the b.u.t.terflies.

But as social att.i.tudes changed, the price fell to reflect the new reality. This shift in preferences has not been confined to prost.i.tution. Among U.S. teenagers, oral s.e.x is on the rise while s.e.xual intercourse and pregnancy have fallen. Some might call it coincidence (or worse), but we call it economics at work.

The lower price for oral s.e.x among prost.i.tutes has been met by strong demand. Here is a breakdown of the market share of each s.e.x act performed by the Chicago prost.i.tutes:

Included in the "other" category are nude dancing, "just talk" (an extremely rare event, observed only a handful of times over more than two thousand transactions), and a variety of acts that are the complete opposite of "just talk," so far out of bounds that they would tax the imagination of even the most creative reader. If nothing else, such acts suggest a prime reason that a prost.i.tution market still thrives despite the availability of free s.e.x: men hire prost.i.tutes to do things a girlfriend or wife would never be willing to do. (It should also be said, however, that some of the most deviant acts in our sample actually include family members, with every conceivable combination of gender and generation.)

Prost.i.tutes do not charge all customers the same price. Black customers, for instance, pay on average about $9 less per trick than white customers, while Hispanic customers are in the middle. Economists have a name for the practice of charging different prices for the same product: price discrimination.

In the business world, it isn't always possible to price-discriminate. At least two conditions must be met:

Some customers must have clearly identifiable traits that place them in the willing-to-pay-more category. (As identifiable traits go, black or white skin is a pretty good one.)The seller must be able to prevent resale of the product, thereby destroying any arbitrage opportunities. (In the case of prost.i.tution, resale is pretty much impossible.)

If these circ.u.mstances can be met, most firms will profit from price discriminating whenever they can. Business travelers know this all too well, because they routinely pay three times more for a last-minute airline ticket than the vacationer in the next seat. Women who pay for a salon haircut know it too, since they pay twice as much as men for what is pretty much the same haircut. Or consider the online health-care catalog Dr. Leonard's, which sells a Barber Magic hair trimmer for $12.99 and, elsewhere on its site, the Barber Magic Trim-a-Pet hair trimmer for $7.99. The two products appear to be identical-but Dr. Leonard seems to think that people will spend more to trim their own hair than their pet's.

How do the Chicago street prost.i.tutes price-discriminate? As Venkatesh learned, they use different pricing strategies for white and black customers. When dealing with blacks, the prost.i.tutes usually name the price outright to discourage any negotiation. (Venkatesh observed that black customers are more likely than whites to haggle-perhaps, he reasoned, because they're more familiar with the neighborhood and therefore know the market better.) When doing business with white customers, meanwhile, the prost.i.tute makes the man name a price, hoping for a generous offer. As evidenced by the black-white price differential in the data, this strategy seems to work pretty well.

Other factors can knock down the price customers pay a Chicago prost.i.tute. For instance:

The drug discount isn't much of a shock considering that most of the prost.i.tutes are drug addicts. The outdoors discount is partially a time discount because tricks performed outdoors tend to be faster. But also, prost.i.tutes charge more for an indoor trick because they usually have to pay for the indoor s.p.a.ce. Some women rent a bedroom in someone's home or keep a mattress in the bas.e.m.e.nt; others use a cheap motel or a dollar store that has closed for the night.

The small discount for condom use is surprising. Even more surprising is how seldom condoms are used: less than 25 percent of the time even when counting only v.a.g.i.n.al and a.n.a.l s.e.x. (New customers were more likely to use condoms than repeat customers; black customers were less likely than others.) A typical Chicago street prost.i.tute could expect to have about 300 instances of unprotected s.e.x a year. The good news, according to earlier research, is that men who use street prost.i.tutes have a surprisingly low rate of HIV infection, less than 3 percent. (The same is not true for male customers who hire male prost.i.tutes; their rate is above 35 percent.)

So a lot of factors influence a prost.i.tute's pricing: the act itself, certain customer characteristics, even the location.

But amazingly, prices at a given location are virtually the same from one prost.i.tute to the next. You might think one woman would charge more than another who is less desirable. But that rarely happens. Why?

The only sensible explanation is that most customers view the women as what economists call perfect subst.i.tutes, or commodities that are easily interchanged. Just as a shopper in a grocery store may see one bunch of bananas as pretty much identical to the rest, the same principle seems to hold true for the men who frequent this market.

One surefire way for a customer to get a big discount is to hire the prost.i.tute directly rather than dealing with a pimp. If he does, he'll get the same s.e.x act for about $16 less.

This estimate is based on data from the prost.i.tutes in Roseland and West Pullman. The two neighborhoods are located next to each other and are similar in most regards. But in West Pullman, the prost.i.tutes used pimps, whereas those in Roseland did not. West Pullman is slightly more residential, which creates community pressure to keep prost.i.tutes off the streets. Roseland, meanwhile, has more street-gang activity. Even though Chicago's gangs don't typically get involved in pimping, they don't want anyone else horning in on their black-market economy.

This key difference allows us to measure the impact of the pimp (henceforth known as the pimpact). But first, here's an important question: how can we be sure the two populations of prost.i.tutes are in fact comparable? Perhaps the prost.i.tutes who work with pimps have different characteristics than the others. Maybe they're savvier or less drug addicted. If that were the case, we'd merely be measuring two different populations of women rather than the pimpact.

But as it happened, many of the women in Venkatesh's study went back and forth between the two neighborhoods, sometimes working with a pimp and sometimes solo. This enabled us to a.n.a.lyze the data in such a way that isolates the pimpact.

As just noted, customers pay about $16 more if they go through a pimp. But the customers who use pimps also tend to buy more expensive services-no manual stimulation for these gents-which further b.u.mps up the women's wages. So even after the pimps take their typical 25 percent commission, the prost.i.tutes earn more money while turning fewer tricks:

The secret to the pimps' success is that they go after a different clientele than the street prost.i.tutes can get on their own. As Venkatesh learned, the pimps in West Pullman spent a lot of their time recruiting customers, mostly white ones, in downtown strip clubs and the riverboat casinos in nearby Indiana.

But as the data show, the pimpact goes well beyond producing higher wages. A prost.i.tute who works with a pimp is less likely to be beaten up by a customer or forced into giving freebies to gang members.

So if you are a street prost.i.tute in Chicago, using a pimp looks to be all upside. Even after paying the commission, you come out ahead on just about every front. If only every agent in every industry provided this kind of value.

Consider a different sales environment: residential real estate. Just as you can sell your body with or without the aid of a pimp, you can sell your house with or without a Realtor. While Realtors charge a much lower commission than the pimps-about 5 percent versus 25 percent-the Realtor's cut is usually in the tens of thousands of dollars for a single sale.

So do Realtors earn their pay?

Three economists recently a.n.a.lyzed home-sales data in Madison, Wisconsin, which has a thriving for-sale-by-owner market (or FSBO, p.r.o.nounced "FIZZ-bo"). This revolves around the website FSBOMadison.com, which charges homeowners $150 to list a house, with no commission when the home is sold. By comparing FSBO sales in Madison with Realtor-sold homes in Madison along several dimensions-price, house and neighborhood characteristics, time on market, and so on-the economists were able to gauge the Realtor's impact (or, in the interest of symmetry, the Rimpact).

What did they find?

The homes sold on FSBOMadison.com typically fetched about the same price as the homes sold by Realtors. That doesn't make the Realtors look very good. Using a Realtor to sell a $400,000 house means paying a commission of about $20,000-versus just $150 to FSBOMadison.com. (Another recent study, meanwhile, found that flat-fee real-estate agents, who typically charge about $500 to list a house, also get about the same price as full-fee Realtors.)

But there are some important caveats. In exchange for the 5 percent commission, someone else does all the work for you. For some home sellers, that's well worth the price. It's also hard to say if the Madison results would hold true in other cities. Furthermore, the study took place during a strong housing market, which probably makes it easier to sell a home yourself. Also, the kind of people who choose to sell their houses without a Realtor may have a better business head to start with. Finally, even though the FSBO homes sold for the same average price as those sold by Realtors, they took twenty days longer to sell. But most people would probably consider it worth $20,000 to live in their old home for an extra twenty days.

A Realtor and a pimp perform the same primary service: marketing your product to potential customers. As this study shows, the Internet is proving to be a pretty powerful subst.i.tute for the Realtor. But if you're trying to sell street prost.i.tution, the Internet isn't very good-not yet, at least-at matching sellers to buyers.

So once you consider the value you get for each of these two agents, it seems clear that a pimp's services are considerably more valuable than a Realtor's. Or, for those who prefer their conclusions rendered mathematically:

PIMPACT > RIMPACT

During Venkatesh's study, six pimps managed the prost.i.tution in West Pullman, and he got to know each of them. They were all men. In the old days, prost.i.tution rings in even the poorest Chicago neighborhoods were usually run by women. But men, attracted by the high wages, eventually took over-yet another example in the long history of men stepping in to outearn women.

These six pimps ranged in age from their early thirties to their late forties and "were doing pretty well," Venkatesh says, making roughly $50,000 a year. Some also held legit jobs-car mechanic or store manager-and most owned their homes. None were drug addicts.

One of their most important roles was handling the police. Venkatesh learned that the pimps had a good working relationship with the police, particularly with one officer, named Charles. When he was new on the beat, Charles hara.s.sed and arrested the pimps. But this backfired. "When you arrest the pimps, there'll just be fighting to replace them," Venkatesh says, "and the violence is worse than the prost.i.tution."

So instead, Charles extracted some compromises. The pimps agreed to stay away from the park when kids were playing there, and to keep the prost.i.tution hidden. In return, the police would leave the pimps alone-and, importantly, they wouldn't arrest the prost.i.tutes either. Over the course of Venkatesh's study, there was only one official arrest of a prost.i.tute in an area controlled by pimps. Of all the advantages a prost.i.tute gained by using a pimp, not getting arrested was one of the biggest.

But you don't necessarily need a pimp to stay out of jail. The average prost.i.tute in Chicago will turn 450 tricks before she is arrested, and only 1 in 10 arrests leads to a prison sentence.

It's not that the police don't know where the prost.i.tutes are. Nor have the police bra.s.s or mayor made a conscious decision to let prost.i.tution thrive. Rather, this is a graphic example of what economists call the princ.i.p.al-agent problem. That's what happens when two parties in a given undertaking seem to have the same incentives but in fact may not.

In this case, you could think of the police chief as the princ.i.p.al. He would like to curtail street prost.i.tution. The cop on the street, meanwhile, is the agent. He may also want to curtail prost.i.tution, at least in theory, but he doesn't have a very strong incentive to actually make arrests. As some officers see it, the prost.i.tutes offer something far more appealing than just another arrest tally: s.e.x.

This shows up loud and clear in Venkatesh's study. Of all the tricks turned by the prost.i.tutes he tracked, roughly 3 percent were freebies given to police officers.

The data don't lie: a Chicago street prost.i.tute is more likely to have s.e.x with a cop than to be arrested by one.

It would be hard to overemphasize how undesirable it is to be a street prost.i.tute-the degradation, the risk of disease, the nearly constant threat of violence.

Nowhere were the conditions as bad as in Washington Park, the third neighborhood in Venkatesh's study, which lies about six miles north of Roseland and West Pullman. It is more economically depressed and less accessible to outsiders, especially whites. The prost.i.tution is centered around four locations: two large apartment buildings, a five-block stretch of busy commercial street, and in the park itself, a 372-acre landmark designed in the 1870s by Frederick Law Olmsted and Calvert Vaux. The prost.i.tutes in Washington Park work without pimps, and they earn the lowest wages of any prost.i.tutes in Venkatesh's study.

This might lead you to think that such women would rather be doing anything else but turning tricks. But one feature of a market economy is that prices tend to find a level whereby even the worst conceivable job is worth doing. As bad off as these women are, they would seem to be worse off without prost.i.tution.