When they got their report cards last year, six of every 10 government managers and supervisors got a job rating that is the federal equivalent of "able to leap tall buildings in a single bound."

This year, under pressure from the top, only about one-third of the 150,000 feds under merit pay -- including about 60,000 Grade 13, 14 and 15 supervisors and managers here -- will get marks that will put them in the running for raises in excess of the 4 percent general federal pay increase.

Under procedures set up by the Carter administration, federal managers and supervisors (in the $34,930-to-$57,500 pay range) are supposed to get performance-linked raises. That system made them ineligible for automatic longevity raises (worth about 3 percent) and guarantees them only half of the regular percentage pay raise most other white-collar feds get each October.

This year's raise for most of the area's 240,000 white-collar rank-and-file workers is 4 percent. Merit pay people received only 2 percent so far. Additional raises will be determined by what kind of ratings they get. Most agencies use a five-tier system -- outstanding, exceeds fully satisfactory, fully satisfactory, minimally satisfactory and unacceptable.

In 1981, 65 percent of the people under merit pay got one of the two top grades. Because of a last-minute funding snafu, everybody under merit pay got the same raise, so the ratings had no financial effect. This year they will.

People rated in the two bottom categories will get 2 percent. Those judged "fully satisfactory" will get something more than the 4 percent increase, depending on the grading curve in their agencies. Those with top marks will get the most.

Most agencies have told graders to be less generous with top marks this year.

In some agencies initial performance ratings have been lowered as they moved up the chain of command for final approval.

Typical of the harder-line approach is the Sept. 23 memo to Housing and Urban Development officials from Secretary Samuel R. Pierce Jr. He said last year's governmentwide ratings for executives, rank-and-file workers and merit pay employes were discussed at a recent Cabinet meeting.

The ratings, Pierce said, "appeared inflated." He said the "concentration of ratings on the high side indicates that managers are not making distinctions among the performance of their employes . . . .

"This year, I want all managers to realize that a rating of Fully Satisfactory should be the norm . . . Higher ratings should be given only when warranted and supported by statements of demonstrable accomplishments which clearly meet the standards for the upper ranges of performance."

Officials at the Office of Personnel Management said that other agencies also are telling top managers to do a better grading job.

The congressional Federal Government Service Task Force says it has heard from employes in several agencies -- one of them the National Institutes of Health -- that performance ratings given them by immediate supervisors have been knocked down when higher officials reviewed them.

An OPM official said, "I see nothing wrong with taking a more realistic look at performance ratings . . . not everybody leaps over tall buildings. A lot of us are very average and that is a tough point to get over to people in the U.S.A." He said it is unfair for officials to give too many high performance ratings, because it cuts down the size of merit pay raises for people "who really deserve them."

As of now, about half of the people under merit pay have been given their initial performance ratings. Everybody will have them by the end of the year. Merit raises approved later will be retroactive to this October.

But because of the new emphasis on hard grading, a lot of people who made the dean's list last year will find themselves rated as C+ this time around.