What to do about Mexico's oil company, Pemex, may seem like a parochial issue of interest only to Mexicans and a few oil industry executives. But the matter should be of concern to anybody who is wondering when oil will come down off its near-record highs.

Pemex generates two fifth's of the Mexican government's income and is a lucrative employer, but it is ailing from neglect. For years the government has milked Pemex of cash without giving it the wherewithal to invest in and develop new sources of oil. When President Felipe Calderon proposed last week to reform Pemex and encourage more private investment in oil exploration and refining, his leftist opponents shut down the country's legislature in protest.  Pemex, they claimed, is a cherished national treasure that must not be pushed into private hands.

Mexico is hardly the only country that treats its state oil companies as ATMs for governments, unions, cronies and others who siphon the rich benefits for themselves. A large fraction of the world's oil patch is struggling with the problem that bedevils Calderon: how to make state-owned oil companies (which control about three quarters of the world's oil reserves) more effective at finding and producing oil. Veneuzuela's oil output is flagging.  Russia's state-owned gas company, Gazprom, is on the edge of a steep decline in production. And in different ways many of the world's state-owned oil companies are are struggling to keep pace with rising demand. Simply privatizing them is politically difficult, and thus most of the world's oil-rich governments are struggling to find ways to make state enterprises perform better.

Even among state oil companies, Pemex's performance is notably poor. Used as a cash cow for the government, Pemex has never been able to keep enough of its profits to invest in exploration and better technology, the lifeblood of the best oil companies. Until a few years ago, Pemex invested essentially nothing in looking for new oil fields. It relied, instead, on the aging Cantarell field, which was discovered in the 1970s not by Pemex but by fisherman who were angry that the seeping oil was fouling their nets and assumed that Pemex was to blame.  Pemex brought the massive field online with relatively simple technology. A scheme in the late 1990s extended the life of the field, but that effort has run out of steam. On the back of Cantarell's decline, total output from Pemex is sliding; some even worry that Mexico could become a net importer of oil in the next decade or two. They're probably wrong, but even the idea makes people nervous.

At times over the last few decades (including today) Pemex has been blessed with a dream team of smart managers, but even they have not been able to reverse the tide of red ink. That's because the company's troubles run so deep that even the best management can't fix them. Indeed, the most striking thing about Calderon's proposed reforms is that they don't go nearly far enough to make Pemex a responsive company, even though they are on the outer edge of what's probably politically feasible in Mexico.

For example, Calderon proposes a new system of "citizen bonds" that will help bring capital to the company (and because they would be owned by the public, these bonds would help blunt the legal block to any reform—Mexico's Constitution requires that its hydrocarbons be owned by the people). Money alone, though, won't reverse Pemex's fortunes. Part of the problem is that risk taking, which is essential to success in oil, is strongly discouraged. My colleagues at Stanford, in a study released last week, have shown that a system of tough laws that control procurement make managers wary of projects that could fail. Although such laws are designed to help stamp out corruption, a noble goal, they are administered by parts of the Mexican government that know little about the risky nature of the oil business.