ONE OF THE MORE peculiar games these days is guessing just how many vacant apartments there are in town - at least that's what's being played in several District government offices. This numbers game got started because local law allows rental apartment buildings to be converted into condominiums if the vacancy rate for the city's lower-income one-bedroom apartments is above a certain level. When the rate is higher than 3 percent, the entire rental market is considered "adequate," and some apartments can be converted-presumably without having too much effect on the total housing supply.

In an attempt to find out just how high -or low-the city's vacancy rate actually is, the Metropolitan Washington Council of Governments, assisted by the city's planning office, conducted a survey of some apartment buildings in town. More than 1,100 questionnaires were mailed to landlords of all types of units, asking about rental cost, size, utility cost and the number of vacancies, if any. A little more than one-half of those questionnaires were returned, providing information on about 53,000 individual apartments. After the answers were tallied and "adjusted," the COG staff members concluded that there is about a 4.12 percent vacancy rate for the entire city-and a slightly higher rate for lower-income one-bedroom apartments.

Just a few days ago, the city's Department of Housing and Community Development held a hearing on the results of the survey and, not too surprisingly, there were a number of questions raised about whether the conclusions were valid. Several of those who testified criticized officials for basing their findings on far too few apartment buildings, possibly underestimating the number of available units in town. Others were concerned that the landlords who responded may have overestimated the number of vacancies, and since they were not visited, there is no way to verify those answers.

Whether those points prove that the survey is invalid is arguable. But there are two other issues that should cause officials to consider starting from scratch on this whole vacancy numbers business. First, none of the landlords was asked if vacant apartments met even the minimum requirements of the city's housing code. That means that even though there may be a rough indication of the number of units that are vacant, there's no guarantee that anyone could live in them. Second, the survey didn't attempt to clarify what "vacant" means-there is no distinction between apartments that are vacant, though there's someone in line ready to move in, and those that are just available , with no tenant in sight.

Determining the city's vacancy rate is no idle exercise; thousands of households will be affected by those numbers, no matter what they are. The COG study is only one step toward getting a true picture of the District's rental market. But Mayor Washington and the City Council will need more information than is now available if they are to figure out a true vacancy rate for the city.