Hats off to Bill Archer. Approaching his final lap as chairman of the Ways and Means Committee, Archer (R-Texas) has dedicated himself to the task of reforming the Social Security system. With the appropriations bills foundering, his party leadership dithering and people talking more about changing Medicare (by adding new benefits) than transforming Social Security, Archer's determination seems, well, quixotic.

Even worse for him, his initial ideas, trying to nudge the system toward privatization without hitting the hot buttons of either party or ideology, have instead attracted the disdain of both liberals and conservatives.

But by refusing to give in to the conventional wisdom that Social Security reform is a dead issue in this Congress, and by signaling that the only real option must be broadly bipartisan, Archer is on the right track. And in fact, his opening has been not rejected but encouraged by his Democratic counterpart, Rep. Charlie Rangel (D-N.Y.), and by the White House and its emissary Gene Sperling. A deal may be doable (more on this below). Let's make sure it is the right kind of deal.

There are compelling reasons to focus on Social Security (and Medicare, for that matter) now. But there is a clear and present danger to be aware of here. Most of the formulas for "reform" are feel-good ones that would add goodies and expand the size of the programs in the future, instead of restraining their future growth. These reforms would be applauded by the elderly and let both parties declare victory. They would also be cures worse than the diseases.

Currently, the two mega-programs of Social Security and Medicare make up about 35 percent of the federal budget, up from about 25 percent when Ronald Reagan became president. In five years, they will be about 40 percent of the budget. Add in Medicaid, which is increasingly a program of long-term care for the elderly, and the number will be more than 50 percent of the budget. With the kinds of changes being considered, they could reach 70 percent or more of the budget by 2030. Add in interest payments and national defense, and that's it.

Without lawmakers noticing it, Social Security and Medicare have become the giant tapeworms of the federal government, crowding out everything else. We are moving toward a zero-sum game where every additional dollar allocated for Social Security will have to come from such other programs as Head Start, embassy security, purchasing errant weapons-grade plutonium, environmental cleanup, health research or education. We are moving toward a federal government that is not spending our collective resources on investments in the future like basic research on health and science or education, but on transferring wealth from one group in society to another, and not particularly from the haves to the have-nots.

Real safety-net protection in retirement income and health insurance and care are and should be top priorities of the nation and the federal government. But reforms in the programs have to achieve those goals without squeezing out everything else.

What to do? Let's start with a little analysis. The projected imbalances in Social Security and Medicare flow from the projected costs of programs that will increase under relentless demographic pressure, and payroll tax revenues that will decline from the opposite side of the demographic coin, i.e., the decline in the work force relative to the retirees.

To solve the imbalance problems, we can find ways to reduce future costs, or find ways to increase future revenues. With a high-stakes election approaching and all eyes on the potent political power of the elderly, few voices are being heard to cut future costs. Most of the suggestions to increase revenues avoid the unpopular (and unwise) suggestion to increase payroll taxes. Instead, the suggestions are to change investments to create higher rates of return, and supplement the program with revenues from projected budget surpluses.

Both parties are falling over themselves and each other to add more surplus monies to save Social Security -which will only add to the problem of an expanded and bloated Social Security crowding out other worthy program areas. Most of the partisan and ideological disagreement has focused on whether and how to change the investment policy of the program, which now takes all payroll tax revenues and puts them into U.S. Treasury paper.

Of course, the debate over changing investments is about more than ways to save Social Security from bankruptcy. Conservatives want to shift Social Security from investing collectively in Treasury paper to investing individually in stock markets to privatize the system and to make everybody in America an investor.

Liberals disdain the complexity and administrative costs of 200 million individual investor accounts, and don't want a radical fix in a system they believe is not broken. President Clinton has tried to bridge this gap by keeping the basic system intact while encouraging an expansion in the investor class through means-related tax credits for saving via his "USA Accounts." But his attempted compromise has been rejected by most Republicans as too little, too costly and just another Democratic tax break.

Changing the investment dynamic won't itself "solve" the Social Security problem. But finding a compromise is a necessary component of broader reform.

Here is a compromise that can begin to achieve worthy goals with a broad coalition of support. It should start with USA Accounts. Conservatives are foolish to reject them outright - they represent a tax cut that benefits poor and working poor Americans, but does so by encouraging people with no savings or investments to save and invest - by expanding sharply the members of the investor class and giving them a piece of the action in the society.

To be sure, USA Accounts are an adjunct to Social Security, not a change; they don't redirect any share of payroll taxes into individual investments or alter in any way the basic Social Security System.

To do that, add in another program that has achieved an astonishingly wide range of support in the political system - Kid Save.

Kid Save has been endorsed by the American Association of Retired Persons and a wide range of liberal and conservative groups. The basic idea is to take a sliver of the payroll tax and use it to invest $1,000 in an individual account for each of the four million Americans born each year.

The investment options would be the same as those for federal retirement accounts. Kid Save would add $500 a year for five years to these accounts, leaving each child with at least $3,500 or more at the age of 5. Through the miracle of compounding, this sum would grow to a quarter-million or more by retirement age.

Under Kid Save, at a relatively low cost, every American eventually would have a robust nest egg, and a piece of America. The impact on Social Security would be delayed, but real.

So we start with USA Accounts and Kid Save. Those are the carrots. Now we need a meaningful stick, to bring some real but fair restraint to the future growth of the program.

The best way to go is to focus on early retirement. Most Americans retire early. They'd be crazy not to - under current law, you can retire at 62 with 80 percent of your full Social Security benefits or wait until 65 for the full amount. Getting most of the benefits three years early makes good sense.

But life expectancy is expanding sharply - the fastest growing age group in America is the over-85s. Retirement at 62 means more Americans getting benefits for more years, adding to the overall burdens on the system. Our policy to discourage everybody from retiring early has been to penalize retirees who earn extra money to supplement their Social Security stipends - a policy under siege from retirees who rightly view a policy to discourage work in a society short of workers as unfair.

But eliminating the penalty for extra earnings would encourage even more workers to retire early and earn Social Security longer. The problem and gap is already likely to worsen.

Starting this year, Social Security's retirement age begins to go up gradually from 65 to 67 under the 1983 reform plan. But early retirement stays at 62. Currently, Americans can retire at 62 with 80 percent of their full Social Security benefit. That number is scheduled to drop gradually under the 1983 reforms, to 60 percent. But the slow shift from 80 to 60 percent, along with the widening gap between early and regular retirement is going to create a bigger fiscal headache.

What to do? Let's combine a removal of any penalty for retirees who work with a gradual shift in early retirement age to 64 or 65, and at the same time gradually lower the early retirement stipend to 50 percent of the full amount. Surveys show that Americans oppose any increase in retirement age - but are much more open to changes in early retirement. This combination of changes makes demographic, social and fiscal sense, and brings common-sense restraint to a program that needs it.

I'm not naive. The chances of any significant movement this Congress to change Social Security are small. But ways do exist to find bipartisan reforms that would reduce rather than exacerbate the problem.