George W. Bush has staked his presidential legacy (and a whole lot more) on a bid to create democracy in Iraq, the centerpiece of his "freedom agenda." But he has made two crucial mistakes. He has raised unreasonably high expectations among Americans for the success of this monumentally complex undertaking, and he has failed to level with the American people about the true cost in blood and resources that such an effort would require. More than three-and-one-half years into the conflict, the president has lost most of the public confidence he enjoyed in 2003.

American presidents have made similar mistakes before-and not so long ago. Following the collapse of the Soviet Union, Clinton Administration officials embarked on a plan to help shepherd the new Russia through "shock therapy" and a series of open elections toward free-market democracy. Expectations for success were high. But a considered long-term U.S. commitment to Russian democratization simply did not exist. Troops were not needed as they are in Iraq. But substantial political and financial support were required-and not forthcoming.

So many of the early visuals were (deceptively) encouraging. Exuberant crowds toppled Saddam's statues and American tanks rolled seemingly unimpeded through the streets of Saddam's capital. And then there were the ink-stained fingers of Iraqis who had braved a variety of threats to exercise their newly acquired right to vote.

But democracy and the open society needed to nourish it requires more than the ouster of the dictator and the holding of peaceful elections. It demands the steady, long-term development of governing institutions that are independent of one another, which trump the power of the country's dominant political personalities and which earn the faith of its citizens.

To build democracy in a state with little or no democratic history is the work of decades-and it can't be done on the cheap. To invest considerable human, political and financial capital in support of the construction of democracy in places like Iraq and Afghanistan simultaneously, as if national elections and good police work will create an inexpensive and self-sustaining momentum toward stable political pluralism, is foolhardy.

But 17 years on from the giddy celebrations atop the Berlin Wall, democracy itself has lost some of its luster. After several years of political, economic and social turmoil in Eastern Europe and the former Soviet Union, large numbers of exhausted voters have opted for stability-with the greater promise of democracy deferred until their governments have inspired greater confidence in official competence. Think of this transition as Yeltsin atop the tank to Putin atop the "security vertical."

To understand why this is so, consider the relationship between a state's stability and its "openness." A country's stability is a measure of its government's capacity to implement policy in the event of a political, social or economic crisis. Openness is a measure of the degree to which people, ideas, information, goods and services flow freely in both directions across the state's borders and within the state itself.

Some countries-Germany, Japan, the United Kingdom or Brazil-are stable precisely because they are open. Commercial, intellectual and social interactions across borders render their cultures and economies ever more dynamic. Other states, such as North Korea, Iran or Cuba, remain stable only so long as they remain closed and isolated.

Now imagine a graph on which the vertical axis measures stability and the horizontal axis measures openness. States that appear as points higher on the graph are more stable, while those lower are less stable. Those further to the right on the graph are more open; those on the left, less so. Were you to map every possible combination of stability and openness that a given state could generate, those points would form a shape very much like the letter "J".

This "J-Curve", properly understood, can be an effective tool for the architects of foreign policy, because it reveals a great deal about how other states define their interests and develop governing strategies. The J-Curve is agnostic on questions of the relative virtues of democracy. There are plenty of states that are more stable as dictatorships than they would be if they were governed with the consent of the governed. The relationship between stability and openness also reveals that a democracy's success depends not on happy accidents of history, but on social, cultural and economic circumstances within a country that do or do not favor its development.

Because the left side of the J-Curve is considerably steeper than the right, it means that whenever a closed society begins to open up and interact with a globalized world, the chances are much higher that such a state can more quickly and easily fall down the curve into instability than a state that is on the right side of the curve. The process of democratization can create instability especially in states where democracy is a recent import, where long-repressed demands for change are released and when previously disenfranchised players scramble for the first time for a share of the country's political and economic spoils.

A state's relative prosperity determines its baseline for stability. If North Korea suddenly struck oil, it would become more stable at every point along the curve, because it would have more resources with which to artificially reinforce stability at every given level of openness. If oil prices suddenly crashed to $20 per barrel, Saudi Arabia would become considerably less stable. But the basic relationship between stability and openness (the shape of the curve) remains the same whether the entire curve is rising or falling.

Given the steepness of the left side of the curve, a formerly closed state that has begun to implement some reforms can more quickly restore damaged stability than a state that has gone further down the reform road-in its attempt to reach the levels of openness and stability of right-hand states. It is faster and easier to restore stability by declaring martial law-closing the country-than it is to create an open, stable, functioning civil society and attract foreign investment. This is especially the case if large amounts of external support are not available to guarantee a baseline level of prosperity while changes are underway or-unlike in the case of a number of Central and East European states-there is no guarantee that the adoption of painful reforms will lead to guaranteed results.

And authoritarian leaders have real incentives to isolate their peoples from the outside world. Not only does any opening in these states, even an incremental one, dramatically raise the prospects of instability, but the continued survival of many of these regimes depends on controlling the flow of people, ideas, information, goods, services and cash into and out of the country, to ensure their position remains unchallenged from within.

Events in Russia over the last 15 years show how states move up and down the curve-and why populations democratically endorse authoritarian measures.

During the chaos of the Yeltsin years, Russians were buffeted by considerable political, economic and social turmoil. Inflation and economic collapse stripped away livelihoods and ruined lives. Beginning in 2000, newly elevated President Vladimir Putin restored Russian stability by concentrating political power in the Kremlin, curbing free expression in the country's media, restricting foreign investment in "strategic sectors of the economy", emasculating civil society and consolidating economic power in the hands of the state. (The tripling of oil prices over the last four years has made his work much easier.)

This forceful re-imposition of order has earned Putin a 70-plus percent approval rating. Broadly speaking, Russians have chosen the order that flows from authoritarianism over the chaos they believe was generated by ill-considered attempts to impose Western-style democracy. This dynamic helps explain why so many ordinary Russians shrug off sermons on the values of liberal democracy-and why isolating the country is not a particularly useful strategy.

The J-Curve also casts light on the current political trajectories in Afghanistan and Iraq. Afghans have again put their trust in warlords-and in some cases the Taliban-because they recognize that authoritarians can more quickly and easily restore stability than the elected (but crippled) central government can. In Iraq, sectarian violence has become a horrifying part of daily life in many areas. When a people faces the daily uncertainties of life in a dangerously unstable country, it values stability above all else. Freedom from fear and insecurity trumps the freedom to vote or to open a newspaper. Until Iraq's Shi‘a, Sunnis and Kurds finally win freedom from fear of sectarian attack and economic exclusion, they will demand stability.

In both cases, the United States can continue to try to safeguard Afghanistan's and Iraq's security until its political leaders forge the compromises needed to begin the long-term process of democratization. But American and allied troops will not remain in both countries indefinitely, because their taxpayers won't allow it.

With each passing week, replacing the unrealistic (democratization) with the possible (support for a regime that can restore stability) begins to look like the Bush Administration's best remaining option for both Iraq and Afghanistan-especially in order to prevent either country from becoming a training ground for Al-Qaeda or, in the case of Iraq, an Iranian satellite.