WHEN THE nation's unemployment insurance system was created in 1935, its main intent was to tide people through short-term periods when they were thrown out of work through no fault of their own. Some 40 years later, the critical question is how to help the system, which is itself in trouble.

In a mere three years, the system has gone deeply into debt, to the tune of $13 billion, in order to continue paying benefits to a record number of jobless workers, and it may be unable to recover fully without remedial action by the federal government.

In fiscal 1973, a so-called normal period, $4.9 billion in unemployment benefits was paid to fewer than 6 million workers. But in fiscal 1976, a record $20 billion was paid to 14 million workers.

Twenty-three state unemployment systems quietly went into the red. So did the national unemployment trust fund. Some states, particularly in the industrialized Northeast, were forced to borrow from federal general revenues to meet their obligations, and several of them doubt their ability to meet the deadlines for repayment.

The system, Sen. Jacob Javitts (R-N.Y.) warned recently in introducing a so-called "fail-safe" financing reform measure, "is threatened with financial collapse in the wake of the longest and deepest economic decline we have experienced since the 1930s."

The bill's co-author, Sen. Harrison A. William Jr. (D-N.Y.), was only slightly less alarmed. The legislation, he said, "is needed to rescue the nation's unemployment insurance system from potential collapse if the economic recovery teeters and slides back into recession at any time in the next several years."

Critics say such statements reflect the extreme situation in two of the hardest-hit states, not in the nation as a whole. However, while dismissing the idea that either federal or state officials would ever commit political suicide by letting the system collapse, many economists acknowledge that another recession in the near future would fundamentally alter the system's character.

The Carter administration has sidestepped the tangle so far. Even the complex welfare reform package refers only glancingly to the system, perhaps in part because the people collecting jobless insurance benefits generally are not the poor, the young or the elderly but the unnoticed middle class.

If one could find a typical recipient of unemployment compensation, the statistics show, he would probably be white, male, between 25 and 64 years old, with annual earnings somewhere between $5,000 and $15,000.

But the system's problems remain. And the answers are hard to find, because the system itself is exceedingly complex.

There are wheel within wheels, and the current financial crisis cannot be considered separately from the system itself. The obvious but often forgotten fact is that there are actually 52 systems - for the 50 states, the District of Columbia and Puerto Rico - with widely varying rules and payments. Each state is competes with its neighbors for investment capital. Each worries about the impact of raising taxes. Each is fearful of losing its share of the business pie. Questions of funding and the size of benefits checks affect not only state and federal budgets but also the costs of doing business. And since the system is fragmented, with as many standards as there are jurisdictions involved, the financial drain on employers, consumers and, indeed, employees, plus the requirements workers must meet to qualify for benefits, vary considerably. Wide Disparities

LITERALLY a stepchild, unemployment insurance was enacted into law as part the Social Security Act of 1935. But, unlike Social Security, the jobless insurance system was dubbed a federal-state partnership - although, in terms of power relationships, a state-federal partnership would allowed to design their own programs and approaches for administering and financing them, the theory being that they were most familiar with local economic conditions.

Meanwhile, the federal government became responsible for establishing a general framework - determining which board groups should be protected, imposing a federal employers' payroll tax credited against state payroll taxes and responding to major recessions with extended benefits and broader coverage. It all became more complex in the recent recession when two supplemental benefit programs were piled on top of the regular 26-week program and the extended 26th-to-39th-week coverage.

Raymond Munts, economist and author of "Policy Development in Unemployment Insurance," has said that, "If one lesson emerges from this history, it is the danger of establishing a decentralized system with no explicit hierarchy of purposes."

Obviously, state control has led to great inconsistency:

A single worker laid off from a $20,000 job draws anything from $63 to $148 a week, depending on whether the state uses a sliding maximum related to the state's average weekly wage for covered employees or a fixed dollar amount.

A single worker can receive more than a married man, depending on what state he is working in. Thus, a married worker with dependents collects from $115 to $174 in the 11 states that offer dependency allowances. A single worker at the same wage but with no dependents gets from $69 to $125.

As little as $150 in earnings is needed to collect the minimum benefits in Puerto Rico ($7 a week), as much as $1,650 or 16 weeks of work in Washington ($17). As little as $2,100 in earnings warrants the maximum benefit in Nebraska ($80); at least $14,300 is needed in West Virginia ($139).

A person who voluntarily quits or is discharged for misconduct cannot collect in about half the states; in others, the penalties range from two to 25 weeks.

A uniform 20 to 26 weeks of coverage is given to all unemployed workers in 10 states: the others consider prior earnings and/or length of time employed.

All but 10 states have a waiting period in which, although the individuals is otherwise entitled to collect, no benefits are paid in order to curb program costs. However, about 10 states pay benefits retroactively.

Benefits are tax-free, but three states (Alaska, Alabama, New Jersey) deduct employee contributions. Should Standards Be Uniform?

A CLOSER LOOK at variations among states shows that a number of sexually discriminatory practices are still on the books, although the denial of benefits solely because of pregnancy has been prohibited by federal law that will become effective in January, 1978. According to a 1975 Labor Department report:

In some states with dependency allowances, a non-working wife is considered a dependent, but th non-working husband remains the breadwinner unless he is totally and permanently disabled.

In some states - and the District of Columbia - a married man is ineligible for 10 weeks if he quits his job to join his spouse; if the roles are reversed, there is no penalty.

Voluntarily quitting to care for children or an ill relative or to sustain a marital relationship will penalize or deprive individuals from collecting in many states.

Some states are more liberal than others in their treatment of illness and emergencies that occur while an individuals is collecting jobless benefits. Eleven have an illness disability clause, providing that individuals remain eligible for compensation checks even if unavailable for work because they get sick or injured - so long as suitable work has not been refused. In other states, a common cold, observing a religious holidays or attending a funeral could mean forfeiting anywhere from a day to an entire week's benefits.

Inevitably, the question of varying state standards gets caught up in the old debate on states' rights. Rep. James C. Corman, a liberal California Democrat and chairman of the House Ways and Means subcommittee on public assistance and unemployment compensation, contends that "unemployment is not a unique state problem - it's a national problem" and that, accordingly, there ought to be uniform standards.

Others argue that disparities among the states are unavoidable because living standards vary from state to state. Rep. William Steiger (R-Wis.) believes minimum benefits standards would unduly inflate benefit amounts in some states. He also holds that the unemployment compensation system is "a federal-state partnership, and it would be disrupted if you control the benefit amounts."

With the exception of the Carter administration, which has no stated view as yet, every President since Eisenhower has opted for a federal minimum benefit standard. All efforts to achieve that have been stymied by Congress.

The AFL-CIO has an even more ambitious long-term goal - to federalize the system in the name of fairness and efficiency. Those who oppose the federal benefit standard argue that it would be the first step toward a single, federally administered program a la Social Security. John Palmer, a senior fellow at the Brookings Institution, contends, however, that a minimum federal standard has growing validity as the country becomes more uniform and wage scales converge.

But it obvious that states must consider the question of competition from other states with lower wage scales. Rep. William Frenzel (R-Minn.), who opposes a single set of federal benefit standards, believes disparities are purely questions for the states to decide and makes his case with an unusual statements: "If New York wants to set their benefits too low, that's their business, and the District of Colombia too high, that's their business." Realistically, the AFL-CIO's Bert Seidman does not see federalization coming "any time soon," and Palmer says, "It's 50 years down the road before we have federal standards."

Last year, Corman tried to get the House to set a federal minimum standard for benefit levels at 50 per cent of an individual's average weekly wage in covered employment. There would have been a cutoff at two-thirds of the state's average weekly wage for proctected workers. Steiger led the battle to defeat the proposal and, not unsurprisingly, it lost.

The debate involved the so-called wage replacement concept - that unemployment compensation should equal at least half a worker's normal pay. The concept is based on the belief that a man looking for a job should not face substantial alteration in his lifestyle - for example, have his mortgage foreclosed or default on a car loan.

Corman, who plans to keep the issue alive, argues that some states do not provide adequate replacement, in the form of unemployment compensation, for a worker's regular take-home pay. On the other hand, a recent study by Mathematica Policy Research Inc., conducted for the Labor Department, found that a majority of all individuals who received federal supplemental benefits during the recession got at least 60 per cent of their prior earnings.

The study took into account the facts that benefits are tax-free and that work-related expenses are no longer a budget factor. But it failed to weigh the expenses of looking for a job - bus fare, suitable interview attire, costs of lunch, baby-sitters and the higher utility bills accrued by staying at home jobless.

The issue is an emotional one and remains unresolved. Finding the Money

STATE CONTROL of the system results not only on broad variations in benefit and eligibility standards but also in the manner in which state programs are financed.

The basic program is financed by both by state and federal employer payroll taxes, which fund permanent, regular and extended benefits programs. These taxes are funneled into a variety of federal and state accounts in the U.S. Treasury. The federal accounts are used to administer the federal part of the program, to finance half the extended benefits program and to provide loans to the states when they have trouble meeting their bills. THe individual state accounts pay regular benefits and the state share of the cost of extended benefits.

But, because state is free to set its own wage base and tax rates, actual employer taxes vary from state to state and even among individual employers. The size of the state's reserves vary as well.

Each state has an experience-rating system, a major factor in determining employer taxes. It is based on the number of employees the individual firm lays off each year, plus other considerations. The rating side-rules vary among the states.

The net result is a vast disparity among tax rates on payrolls. In Massachusetts last year, the tax rate was 2.5 per cent of taxable wages. In Texas, the rate was 0.6 per cent. Not unnaturally, the two states differed in payments made. The maximum weekly benefit in Massachusetts was $162, in Texas $63.

All this has a bearing on the key question: how to wipe out the massive debts which some states piled up during the recession. A recent report prepared by the Senate Budget Committee at the request of the Labor Department argued that increasing payroll taxes to pay off the debts "may add to both unemployment and inflation" and proposed instead the emergency transfer of credits from general Treasury Department funds to various social insurance accounts.

But the report also concluded that such a proposal was bound to be controversial. The report suggested that state control advocates would view use of the federal treasury to maintain the system's solvency as a step toward federalization. It also noted that states which remained solvent during the recession would object to debt relief for states with heavy deficts.

The total state debt is high - $4.6 billion. (The remaining $8.4 billion in red ink represents are federal unemployment trust fund deficit.) Three states alone - Pennsylvania, New Jersey and Illinois - account for $2.2 billion of the overall state debt.

With the economy improving, states are supposed to pay back their debts or risk stiff federal penalties. But some state administrators argue that, given the precarious condition of their systems and the continued necessity to borrow just to keep pace with claims being filed, imposition of penalty for late payments is a ludicrous idea.

Michigan, for example, is number four on the borrowing list, in debt $624 million to the federal government. And the state's director of employment security, S. Martin Taylor, warns that "a snowball effect" is not inconceivable. That, he says, wwould mean the most vulnerable states would lose industry and pile up higher unemployment rates if they tried to raise revenue to pay back their debts. With fewer jobs and reduced industry, state revenues would fall. And, so, there would have to be additional borrowing, further debt. For an extended period, th payroll tax would jump, with benefits and coverage declining.

Last May, a group of state unemployment insurance administrators, gubernatorial aides, Labor Department officials and congressional staffers met under the sponsorship of the National Governors Conference to consider strategies for securing relief. At first, some states had hope their debts would be "forgiven" by the federal government. But they soon realized this was wishful thinking.

The conference, instead, chose to work for a cost equalization/reinsurance plan which would effectively pool excess benefit costs incurred by depleted state unemployment funds. But it could not resolve the question of who could pay: Should it be the employers of fiscally sound states? Or should the money come from the federal general revenues?

States such as Texas, Georgia and Nevada, which had stayed on a sound financial footing during the recession, fear they will be forced to foot the bill and want to know what they can get out of the deal. There is every indication that they may fight any such plan, particulary since they beleive it was liberal eligibility and benefit standards, plus inadequate trust fund reserve ratios, which got the hard-hit states into the mess to begin with.

But proponents of cost equalization/reinsurance, such as Massachusetts' John Crosier, president of the Interstate Conference of Employment Security Agencies (ICESA), contend that even fiscally stable states could derive benefit.

"You're buying into an insurance program which you might need someday," he told the conference. "The time crunch is so severe we should run with what we've got and get a push to take on Congress."

The push has begun, but with little momentum. When two cost equalization-reinsurance measures were introduced in the House and Senate last month, almost no one noticed.

The ICESA-supported bill, introduced in the House by Rep. William M. Broadhead (D-Mich.), along with 31 co-sponsors, would establish a program to subsidize states experiencing high rates of unemployment through general Treasury revenues.

States with an annual insured unemployment rate of at least 6 per cent would be eligible for repayments of between 25 per cent to 50 per cent of their excess benefit costs. A retroactive provision dating back to 1974 would assist hard-hit states in repaying their debts to the federal trust fund.

To the surprise of some observers, the bill has the guarded support of UBA, an employers' interest group. But the group is disappointed that the bill does not include a provision calling for the federal government to underwrite the cost of the recession's "federal suplemental benefits," the third tier of coverage, rather than require the states to repay advances from the U.S. Treasury totaling $5.8 billion.

"Federal supplemental benefits bear little or no relationship to prior earnings and are a quasi-form of welfare," Eldred Hill, UBA executive director, maintains. "We object to that kind of payment out of trust funds."

The bill introduced in the Senate by Sens. Javits, Williams and six co-sponsors from the Northern industrial states is broader and more liberal. It would:

Establish a cost equalization program similar to the House plan, but provide more generous federal payments - between 50 and 70 per cent of excess costs.

Revise the federal supplemental benefits program by modifying the financial methods and the triggers which put the program in motion as well as the program's duration.

Provide a permanent "fail-safe" program to go into effect in times of crisis, effectively allow states to repay their jobless insurance debts over a five-year period and cut the penalties for late payment. To discourage states from taking advantage of interest-free federal credit, however, the bill would provide for 5 per cent interest on such loans. (A current loophole allows states to repay their loans one day and take out new loans the next.)

Williams speculates that the system could not bear the pressure of another recession under present circumstances. The extra cost of recession-caused joblessness in 1974-76, he says, was $14.5 billion higher in the aggregate than "in more normal times." What of the Future?

WHAT IF the system collapsed? Would the federal government ever allow that to happen? Would it continue letting states use the band-aid, quickie loan approach? No one really seems to have the answer.

Last October, Congress revised the jobless insurance law - expanding coverage to about 90 per cent of the work force and raising employers federal payroll taxes - and established a 13-member National Commission on Unemployment Insurance. The commission was to study the whole system, provide some clues to the answers of questions burdening it and report its findings by July, 1979. It does not yet exist in anything but name.

A chairman has not yet been designated, the commission's composition was challenged as unbalanced by the Labor Department and recommendations on alternate members sent to President Carter earlier this summer have not been acted upon.

At issue are the seven "eleventh hour" appointments made by President Ford prior to his capture from office. (The remaining six appointments were made by the Senate and House.) In addition, critics charge that the commission is heavily weighted against women, labor and public representatives.

"I don't think organized labor would be too enchanted with the composition," Chuck Knapp, special assistant to Labor Secretary F. Ray Marshall, said. "It puts the administration into a difficult position and is a ticklish matter involving a monumental numbers problem."

Knapp conceded it could be another three months before the commission gets off the ground. A White House aide says no action is expected before September.

Just where President Carter fits in the picture remains cloudy - he has never publicly commented on the basic questions of unemployment insurance. The issue has been so seemingly unimportant that, during the 1976 campaign, the only mention made in the Democratic Party platform stated: "Clearly, useful public jobs are far superior to welfare and unemployment payments."

Clearly, too, the administration is in no hurry. Labor Under Secretary Robert Brown doubts the administration will be able to face the House Ways and Means Committee on the financing issue before next year.

The Carter welfare reform package would leave the regular unemployment compensation program intact but would redefine the "triggers" so that, if the national unemployment rate dropped before a specific level, no state could pay "extended" benefits from the 26th through 39th weeks of joblessness regardless of that state's own unemployment burden.

In addition, cash assistance through welfare would be reduced by 80 cents per dollar for those receiving both welfare and jobless benefits. Thus, if a worker received $50 in unemployment insurance, his public assistance would be reduced by $40. The change is justified by the administration on the basis of the welfare package's projected improved cash benefits.

Whatever the value of the welfare program, its proposals hardly touch the gut-level problems faced by average recipients of unemployment compensation. Nor does it address fundamental benefits and financing questions of the jobless insurance network of programs.

The blue-collar man laid off by a stalled assembly line, the white-collar worker who must scout for limited openings while collecting unemployment compensation, the woman who must quit a job to join a spouse in another state - none of these will derive benefits from the welfare reform package. It would also have no impact on existing or potential debts burdening the system or the inequities which so often embitter the jobless.

But, at best, unemployment at a level of at least 5 per cent is going to be with us for a long time to come. That hard look at unemployment compensation, the first in 40 years, is overdue.