Next January, the FICA tax rate (for Social Security and Medicare) is scheduled to rise from 7.15 percent to 7.51 percent, for both employers and employees. It is supposed to rise again in 1990 -- to 7.65 percent. For an employee with the maximum taxable earnings in 1987 ($43,800) who continues at this level in 1988, the tax increase is $157.68; for an average worker it is about $70. The increased revenues will go entirely to Social Security.

Are these increases desirable or necessary? The answer is clearly "no."

Five years ago a financial crisis confronted the Social Security program. The trust fund, which pays retirement and survivor benefits, would not have had sufficient money to pay benefits on time late in 1982 if it had not received loans from the disability and hospital funds. And even these loans provided funding for only the next eight months.

Social Security amendments of 1983 turned the situation around. The fund balance for the two Social Security trust funds has risen steadily. In fact, at the end of 1986, the balance was about $47 billion (approximately three months' benefits).

In part, solvency was restored (and ensured) by tax-rate increases. The increase previously scheduled for 1985 was advanced to 1984, and 72 percent of the increase scheduled for 1990 was advanced to 1988 (with the remainder left for 1990).

What is estimated to occur under this accelerated tax schedule? The estimated fund balance, expressed in 1986 dollars (to remove the effects of inflation) rises rapidly and steadily. It reaches the almost inconceivable height of $2.5 trillion in the early 2020s, about 35 years from now. (In dollars of that time, this would be about $11 trillion.) But after this peak, a rapid decline occurs, and the fund is exhausted in 2051.

This is certainly not a reasonable or logical way to finance a pension plan of any sort. If a large fund is to be built up, so as to provide investment income to help finance anticipated higher future costs, it should not eventually be dissipated.

What happens if the same benefit structure is to be maintained after the fund is exhausted (or without the fund buildup)? The ultimate employer and employee tax rates would have to be increased by about 1.3 percent each over what is now scheduled for 1990 -- not an unmanageable rise.

Should the near-term tax-rate increases be retained and further increases scheduled in the future so that a large fund is built up and is maintained for all time to come? Such a course of action is, in my view, very undesirable. One danger is that the huge balances apparently available (or to be available) would cause irresistible political pressure to liberalize the benefits now -- which would only compound the cost problems some decades hence. Also, the steady and ready availability of large sums for investment in government bonds could well cause increased, unnecessary governmental spending for other purposes, because there would be less need for the federal government to go to the open market for loans. Some might also argue that large Social Security trust-fund balances would be used to balance the general budget.

I believe Social Security should be financed on close to a current-cost basis. Income should slightly exceed outgo each year, in order to build up a fund that is about equal to one year's outgo -- and certainly no more. This should be accomplished by changing the future tax-rate schedule so as to more nearly match the trend of outgo. In the near future, the tax rate should be a little higher than this, so as to build up the fund balance to the desired goal of one year's outgo.

I would freeze the Social Security tax rate at its current level, and then decrease it by 0.7 percent in 1996, after an adequate fund had been accumulated. The tax rate would need to be increased in 2015, and then again in 2020 and 2025.

My proposed tax rates would be lower than those scheduled in 1988 through 2019, higher in 2020-50 and the same thereafter. From 2019 to 2050, the rate would rise from 6.0 to 7.5 percent. These rates are no higher than what would ultimately result under present law if the benefit structure were left unchanged.

Under my proposal, the trust-fund balance would slowly but steadily build up over the years. It would reach $450 billion in 2020 (in 1986 dollars), compared with $12.5 trillion under present law -- and compared with $50 billion today. Then, in 2050, the balance under my proposal would be about $900 billion, as against bankruptcy under present law. The roller coaster effect estimated under present law would be replaced by one of slow but steady growth.

Experience in the future may not follow the intermediate estimate, and then what of my proposal? The answer is that the tax schedule would have to be reviewed from time to time as the experience unfolds, and new estimates of the future experience are prepared. Congress could then legislate different scheduled tax rates for the future. But this would also have to be done if the present tax schedule and funding approach were to be continued.

Congress should reexamine the long-range financing of the Social Security program, and it should do so soon.

The writer has served as chief actuary and deputy commissioner of the Social Security Administration.