You're forgiven if you skipped the Bush administration's plan to overhaul the banking system last week. Let's face it: bank reform is deadly dull and hugely complex. But pay attention anyway. It matters. Sooner or later, something like the Bush proposal -- which expands bank powers and toughens some federal regulation -- is needed to strengthen banks and protect the federal deposit insurance fund from massive losses.

It's no quick fix. Even if Congress instantly adopted the plan (and that won't happen), it wouldn't rescue many already ailing banks. Although most of the nation's 12,400 banks are healthy, roughly 1,000 -- including some major banks -- are not. Their problems mostly reflect poor management: they made a lot of bum loans. But outdated government regulation allowed the troubles to occur.

Historically, each of our major bank reforms has repaired the failure of the previous reform. In the Great Depression, Congress created the Federal Deposit Insurance Corp. (initial coverage: $5,000) to prevent bank panics. In fact, the Federal Reserve had been created in 1913 to do precisely that by providing besieged banks with ample cash to calm nervous depositors. But the Fed couldn't cope. Between 1930 and 1933, two fifths of the nation's 23,679 banks failed.

Unfortunately, yesterday's solution is today's problem. Every banking system requires two things. The first is confidence. If depositors lack it, banks can't count on stable funds with which to make loans. The second is discipline. Without it, bank managers may waste funds on stupid or speculative loans. The trouble is that the steps Congress took in the 1930s to instill confidence have eroded bankers' discipline.

By shielding depositors from bank mistakes, the FDIC paradoxically provides banks with funds to make mistakes. Depositors don't discriminate. They don't favor prudent banks or shun highfliers. This was a major cause of the savings and loan debacle.

What compounds the banks' problem is the loss of their commanding position in the financial system. Between 1950 and 1989, U.S. banks' share of all loans and investments slipped from 50 to 24 percent. Some of the banks' best customers now go elsewhere. For example, many blue-chip companies borrow in the commercial paper market rather than rely on bank loans. To fill the gap, some banks turned to riskier borrowers: developing countries in the 1970s and real-estate developers in the 1980s. In both cases, huge loan losses resulted.

The dilemma of reform is how to restore financial discipline without destroying depositor confidence. The administration's plan -- crafted by Treasury Secretary Nicholas Brady -- errs on the side of maintaining confidence. It hardly touches deposit insurance. The plan would limit any depositor to no more than two insured accounts at any one bank (a regular account and a retirement account). But people could still have as many insured accounts as they wanted by spreading them among many banks.

A trickier issue involves the "too big to fail" doctrine. In theory, the FDIC insurance covers only three quarters of banks' deposits. The rest exceed the $100,000 cutoff. Actually, most deposits are protected. The "uninsured" deposits are concentrated at major banks. When these banks fail, the FDIC usually protects all depositors out of fear that inflicting losses might trigger mass withdrawals of big deposits at other banks -- and thereby create a crisis. Brady's plan would perpetuate this practice.

To strengthen banks, he offers a bargain: more powers in exchange for stricter regulation. Banks could branch across state lines. Some bank holding companies could sell mutual funds, write insurance and deal in stocks and bonds. Meanwhile, banks would face penalties -- restrictions on dividends and growth -- if they fell below minimum capital requirements, to be 8 percent by 1993. (Capital is shareholders' and outside investors' money.) If capital fell too low, banks would be closed before losses had to be made up by the FDIC. Now banks aren't shut until all capital is exhausted.

If banks aren't profitable, they won't be sound. Although many banks operate across state lines, they typically must have separate corporations in each state. The Brady plan would allow this costly system to be scrapped. Duplicative overhead -- computers, staff -- could be cut. Interstate mergers would also be encouraged. Lowell Bryan, a consultant with McKinsey & Co., thinks cost savings could ultimately total $10 billion to $15 billion annually. If true, that would help bank profits, which totaled $15.7 billion in 1989.

Details are crucial. Bryan and some others think banks should be split into two parts. One would qualify for FDIC insurance and would be restricted to less risky lending. In Bryan's scheme, it could invest in U.S. Treasury securities, consumer loans and medium-size business loans. All other loans and investments would be made by a separate entity, where it would be made clear -- perhaps by a minimum deposit of $10,000 or more -- that deposit insurance and "too big to fail" didn't apply. Presumably, the insured bank would pay lower interest rates on deposits.

Congress will debate these and other details. It will also worry about the FDIC's ability to cover the losses of already weak banks. The agency claims that higher payments from healthy banks (which finance the FDIC) can avert a taxpayer bailout. But these matters shouldn't stalemate the debate over long-term reform. Banks need both more freedom and more regulation. A slow fix is better than no fix at all.