"A RBITRON WAS never intended to be used as a sales tool," says Andy "AOkershausen, exective vice president at Washington radio station WMAL. "It was a programming tool to tell a station how its programming was going, but it's turned into a bible for certain people buying broadcast. We're slaves to a system that is not what it was intended to be."

Ratings rule the radio roost in America. They affect what and whom the public listens to, and when and where the public listens. They cause format changes, they raise and lower advertising income, and they enable radio stations and advertisers to pinpoint audiences by age and income.

And Arbitron rules the ratings roost. A good Arbitron report has become equivalent to a good report card for a radio station. Since ratings determine advertising rates, a good report can mean immense profits. Enough bad reports and you're likely to hear a format change.

Says one successful station manager about ratings: "The stations that say they don't need 'em don't have 'em."

As Nielsen rates television, so Arbitron rates radio, by identifying listeners according to age, sex, socioeconomic profile, buying patterns and time-of-day listening habits. There are alternative services, but that's all they are--alternatives. In Washington, Arbitron recently rocked radio by weighting its rankings more heavily toward the black audience. The ensuing controversy was nothing new for a service whose rankings have sometimes appeared to be ranklings.

"Without a doubt Arbitron owns the radio measurement market," says one Washington ad agency rep. With radio rating much more difficult than its television counterpart (there are so many more stations in each market, so many more radio units--five to the average household--and so much listening done outside the home), Arbitron pioneered the use of sophisticated demographic data through an elaborate diary system that replaced the face-to-face interviews and phone method of previous systems. Its client list grew from 475 stations in 1970 to 2,300 in 1982.

In Washington, the stakes in the Arbitron ratings game are high.

The audience is large (Washington is the ninth largest market in the country), with a potential audience of 2,610,200 (over the age of 12) and an effective buying income of $45 billion dollars. In an average hour, 1,936,800 of these potential consumers listen to radio for at least 20 minutes. The market also has a plethora of formats--all-news, all-talk, rock, religious, big band and urban contemporary (black). And, ad agency reps like to point out, Washington is so upscale that there are lots of stations with high "qualitative" audiences.

Advertisers, who ultimately control what we hear by throwing their dollars behind stations that attract the biggest audiences, not only love, but demand, tight demographics that deliver an easily definable audience. In Washington, where a single ratings point can drop a station's position and hence its advertising pull, Arbitron is overwhelmingly accepted, but often grudgingly so.

Some local stations, including WPGC, WHFS and WYCB, refuse to subscribe to it, and almost everyone complains about the costs. There's grumbling that the diary system favors high-power stations and networks; that it overemphasizes older, more literate, upscale listeners; that Arbitron, by not measuring group quarters (colleges, hotels, barracks) doesn't accurately measure the key 18-24 age group, whose members are often reluctant to fill out diaries (therefore overweighting middle-of-the-road and news formats, whose audience are older and more responsive to diaries). The phone-research-based Birch Report is the only significant alternative, but it has neither the scope nor the influence of Arbitron.

But Goff Lebhar, station manager at DC-101, says, "People who bill $7 million or $10 million based on Arbitron are happy. I'm sure it's the best tool today to use for measuring who's listening to radio. But I don't think time should be bought based on Arbitron or any other survey method. As it is now, 99 percent of the buys are made based on something printed somewhere. We'll break the trend when ad agencies and clients go back to 20, 25 years ago when they understood radio and didn't have to go strictly by research numbers."

What Arbitron doesn't measure is a station's total sphere of influence, the attitudes of listeners toward a station, its sense of community or responsibility, its prestige or impact. But it does measure, even within the accepted limitations of all statistical analysis, the thing advertisers understand best: numbers.

"You can sell advertising without ratings," says one station manager. "But to sell lots of advertising, you must show up in the ratings."

Arbitron, created in 1949 as the television-based American Research Bureau, moved into radio in the 1960s. In 1970, it was taken over by computer processing giant Control Data Corp., and changed its name to Arbitron in 1973 after people kept mistaking ARC for a government agency. The firm still does some television and cable research, but its emphasis is radio. "Suffice to say, we're a profitable organization," says Shelley Gagner of Arbitron.

In the fall 1982 ratings book, 29 local stations, accounting for 90.1 of the area's listening audience, were listed. Twenty stations bought the fall book, with "subscriptions" ranging from $35,000 a year paid by WEZR to $127,000 by WKYS/WRC and a top of $152,000 paid by WMAL/WRQX (Q107). The subscription price, says Gagner, depends on "how many times a market is surveyed, how much it costs for us to survey that market, the size of the market, the advertising revenue in the market, how fast the market is growing, how many subscribers."

The books are sent to 3,500 advertisers and ad agencies who use them to pursue specific demographic targets while the stations use their current ratings to try to charge the highest possible rate per spot.

"Remember, whether it's radio or TV, your cost per spot is what the market will bear," said Mike Heimberg of the Capital Media agency.

That enables upscale WMAL (fewer than 2 percent of its listeners are under 18 while half are over 45) to charge as much as $500 a minute while low-rated WEAM can get only $20.

The Arbitron surveys are typical of statistical research. According to Gagner, 4,135 individuals in the Washington area were sent fall 1982 diaries: "Out of that amount we got back--in a usable condition--2,022, a 48 percent response rate, which is about right. We send out and contact about twice what we need."

The diary keepers are "a representative sampling of a population--the keyword is representative--in the same proportions as the makeup of the population. We buy a list of listed telephones and addresses from a firm called Metromail, a company that takes apart phone books all over the country. Once we know the listed phones in an area through a process called Expanded Sample Frame, we can determine the possible unlisted phone numbers; those people have to be included in our sample, because through our research, we found that people with unlisted phones listen differently, they're determined to behave differently. Then we combine the two lists."

"The only people we can't reach are those without phones and we do not measure group quarters--hotels, barracks, colleges--because of the phone problem. But we're working on it.

"Everyone in the home over the age of 12 gets their own little diary, a log of what you're listening to," Gagner says. "Every time you change the station you write down what you turn to, and there's space for comments and demographic information."

Arbitron provides payment-- it calls it an incentive or premium--with the blank diaries. It's not much: 50 cents to $2 a diary.

Homes with at least one black male 18-34 get a higher premium, anywhere up to $5 a person in the home, because it was felt that blacks were not participating in the diary keeping as much as whites.

As a result of this "Differential Survey Treatment," as it's known, there was "an immediate effect on our numbers and conversely on the black stations, which all showed a dramatic increase when Arbitron changed the system," said WMAL's Andy Ockershausen.

The change in methodology--Ockershausen calls it "mythology"--may have been a reason why WKYS has taken over from WMAL at the top of the ratings.

On the other side of the fence, Bart Walsh, general manager at WKYS, insists that "until DST was implemented by Arbitron, they were undersampling certain areas and ethnic groups in their methodology."

The diaries are kept for only a week. The information is then fed into computers. Only 20 percent of the survey information makes it into the Arbitron books.

The rest is available for a fee, and includes more specific breakdowns on age groups, time periods and ethnicity.

Arbitron has other analytical tools, psychographics and socioeconomic profiles, such as PRIZM, a concept of the Rosslyn-based Claritas Corp. PRIZM defines life-style clusters by ZIP codes on the basis that people of the same behavior patterns tend to live near each other. There are 40 different types of code areas, with names such as "Archie Bunker's Neighbors," "Social Climbers" and "Bluebood Estates."

One media buyer pointed out: "Ten years ago, people depended much more on judgment. If you live in a market, you pretty much know what stations people listen to and where they live. There are always buyers who buy by the numbers and don't use qualitative judgment. What separates the good and bad buyer is the one who looks at everything."

But the bottom line in this $75 million market, like it or not: If you rate, you rake.