DESPITE the clamor of debate over school quality, almost nobody is happy with the government's methods for measuring that quality -- especially now that the attention has swung from measures of spending on schools to measures of what students know when they finish. But nobody is making much progress either on developing better yardsticks.

This spring the education secretary's annual "wall chart" ritual highlighted this awkward stalemate. It attempted to compare students' and states' performances over time, but the only figures it could draw on were the notoriously enigmatic SAT and ACT scores -- which measure a different group in every state -- and graduation rates. How is reform to proceed without a better sense of where the problems are? In fact, the government already administers a test much more suited to this purpose, the National Assessment of Educational Progress, and has been planning to expand it from a national test to one that would give every state's scores. But this project is now stalled by a long-standing uneasiness in Congress over standardized tests in general and state-by-state comparisons in particular.

The president's state-of-the-union list of six educational goals included an endorsement of testing at grades four, eight and 12 -- an explicit reference to the NAEP, which now tests a national sample of students in those grades on a rotating list of subjects every two years. In 1969, when Congress created the NAEP, fear of national control -- or a "national curriculum" -- was so deep-seated that the law explicitly barred any breakdowns of scores into states. The "national curriculum" idea has lost some of its unpopularity as more educators accept the notion that the choice is really between a bad "national curriculum" -- determined by textbook companies and inferior tests -- and a better one. But the more urgent argument is that to attack education's problems, states must first describe them far more precisely.

In 1988 Congress voted to try voluntary state-by-state scoring -- once in 1990, again in 1992 -- for eighth-graders only. Thirty-seven states signed on, but it will be 1991 before the first results come through -- these things move glacially -- and after that the law lapses. Continuing past 1992 would require starting almost immediately. Legislation to do so has not been forthcoming, and the White House and the national governors' commission -- bodies calling vociferously for "goals" -- have been silent.

Testing, of course, always embarrasses somebody. Many states fear that low statewide test scores, like high taxes, will scare away industries looking to build. But the inconvenient fact remains that we will never be able to do anything for the schools until we figure out exactly what is going on there, and if anyone is serious about meeting goals "by the year 2000," the time to start was yesterday.