V iolent crime rates in America have declined from coast to coast since 1994. As measured by the Bureau of Justice Statistics (BJS) through its annual household surveys, the total rate of victimizations from violent crime–homicide, rape, robbery, and assault–per 1,000 persons age 12 or older fell by 53.5 percent from 1995 to 2005. With a slight stall in 2006, the declines continued through 2009; as tallied by the FBI from reports filed by some 17,000 state and local law enforcement agencies, the number of violent crimes fell by 10 percent and the number of murders fell by 4.4 percent between 2008 and 2009.

As it has been for much of the last 16 years, the latest good news on crime was led by New York City. The Big Apple’s fabled police force started collecting largely reliable statistics on violent crimes and murders in 1963. In 1990, the city had a record 2,245 known homicides; by 2007, the body count had dropped to a post-1963 record low of 496. After trending up to 523 murders in 2008, the city counted 461 at the end of December 2009, a new post-1963 low. Citywide rates of most other categories of violent crime also declined significantly.

Both in New York City and nationally, the 2009 drops in lethal violence and other violent crimes were all the more impressive given that, in 2008, many criminologists fretted that bad economic conditions, financially strapped and shrunken police forces (for example, in 2009 New York City’s police force was down by about 6,000 officers compared with 2001), and other factors would conspire to put crime back on the rise.

Still, even as we enter a new decade with crime at lower levels than anyone could have reasonably hoped a decade ago, all the good news about crime is not quite as good as it seems. Four sets of facts mug the conventional post-1994 crime-drop narrative.

First, criminologists are far from understanding why crime declined or how to keep crime on the run. Second, the federal government systematically undercounts crime nationally, and does so in ways that confound policy-relevant, city-to-city comparisons. Third, although violence by and against juveniles has receded from record highs (not reached new highs, as had been widely predicted) since 1994, far too many young Americans are still being lost to what Barack Obama has eloquently called "an epidemic of violence that’s sickening the soul of this nation."

Finally, America in 2010 is a half-century into massive public and private anti-crime investments, tougher criminal sentencing policies, and security-seeking personal life style changes; yet, for all that, nationally, crime is not demonstrably lower today than it was 50 years ago, and in many places it is worse than it was when the national alarm about crime was first sounded. Since crime climbed on to the federal policy agenda in the early 1960s, successive government wars on crime and drugs have sacrificed once-sacred civil liberties. Today, nearly 2.5 million Americans live behind prison gates while tens of millions more live in gated communities. We reflexively practice crime-avoidance behaviors that forsake personal freedom to live as and where we would truly like. We spend billions a year on a private-security industry that profits from our crime fears, and we spend billions more on a recession-proof prison-industrial complex. Measured by the number of persons sentenced to a year or more in state or federal prisons per 100,000 citizens, the imprisonment rate was about 100 in 1960 and still under 150 in 1980, but it broke 200 in 1985, 300 in 1991, 400 in 1995, and 500 in 2006. More than 6 million still-living Americans have spent a year or more behind prison bars. Yet we are not safer–and do not feel safer–than our grandparents did a half-century ago.

The post-1994 crime decline furnishes no basis for complacency about crime. Instead, we need government, led by Washington, to get going on fresh crime policies. The Obama-Biden White House should lead on a new federal crime bill before Republican presidential hopefuls or the GOP minions in Congress beat Democrats to it. The interlocking goals should be to promote greater public safety without further draining the public’s purse; to make citizens feel less inclined to lead security-alarmed life styles; and to change federal, state, and local sentencing policies so that, by 2020, civil liberties are more widely respected and America is completing its first post-1960 decade of zero prison growth.

Whither the Crime Drop?

Over the last decade, the national murder rate per 100,000 residents has been between five and six, and New York City’s murder rate has been in that range. What’s true for the Big Apple, however, is far from true for many other cities. New Orleans, for example, recorded 179 murders in 2008 and 173 in 2009 (the city’s per capita murder rate per 100,000 residents in each of the last two years was 50 to 55 if you use the city’s post-Katrina population estimates or 60 to 65 if you follow the FBI’s post-Katrina population estimates). In either case, the New Orleans murder rate is at least ten times the murder rate nationally. St. Louis, Philadelphia, Detroit, and Baltimore, among other cities, have all had some share in the post-1994 decline in crime. But, more similar to the Big Easy than to the Big Apple, their respective crime drops have been neither steep nor stable. Baltimore recorded 130 fewer shootings but four more murders (238) in 2009 than it had in 2008 (234).

Why is the Big Easy so much more crime-ridden than the Big Apple? Why have diverse cities been post-1994 crime-drop laggards? Why are certain neighborhoods in any given city far less crime-ridden than nearby neighborhoods in the same city? And why do national and local crime rates fluctuate over time as they do?

Despite cross-pollination with more robust social sciences like economics, criminologists have not been terribly successful in explaining variance in crime rates. Moreover, they have been markedly unsuccessful in predicting crime trends: Few forecast the crime waves that started in the mid-1960s and the mid-1980s, and many forecast a crime rise that instead became a crime retreat in the mid-1990s. Even retrospective efforts to explain crime patterns and trends rarely result in empirically bankable explanations.

The best book yet on crime’s post-1994 decline is the National Academy of Sciences-sponsored volume, Understanding Crime Trends: Workshop Report, published in 2008. The instantly acclaimed report was produced by a methodological and ideological cross-section of the nation’s most widely respected criminologists. In its single most notable section, Eric P. Baumer persuasively summarizes the evidence on five factors that figured significantly in the crime decline and the estimated range of the crime-drop difference for which each factor was likely responsible: increased incarceration rates (10 to 35 percent); improved economic conditions (10 to 30 percent); decline in teenage births (10 to 35 percent); larger adult cohorts (4 to 8 percent); and increased police per capita (3 to 7 percent).

But there are some immediately obvious points of concern. For one, if all the estimates of the outer-bound impact of the five leading factors were correct, they would explain an impossible "115 percent" of the crime drop. And then, in addition to these five factors, there are other much-discussed factors like the decline in crack cocaine use, enhanced gun-control enforcement, the spread of community-based policing, the 1970s abortion legalization, and the proliferation of "shall-issue" concealed weapons laws. By some estimates published in leading academic journals, various combinations of these factors alone can account for the entire crime decline.

The crime-drop criminology is cloudiest regarding inter-city, intra-city, or year-to-year local fluctuations in crime. For instance, some experts insist that policing mattered a lot to the national crime drop, but others argue that it mattered hardly at all, or only in a few big cities. Some argue that changes in policing tactics made a positive difference, but others conclude that it was more cops per capita, not "community policing" or "broken-windows policing," that made a difference. For every refereed academic journal article reaching one of these conclusions, there is another that reaches the opposite.

Even if we knew what determined the crime drop, we would not therefore know how to engineer factors such as improved economic conditions or fewer teenagers having children. And, even if we could engineer them, what worked then would not be certain to work now. For instance, most experts cite increased incarceration as a major factor in the crime drop, responsible for as much as a third of the total decline. In "Two Million Prisoners Are Enough," a much-debated Wall Street Journal op-ed published in 1999, I popularized the early evidence indicating that post-2000 increases in incarceration would have diminished crime-reduction returns. In Prison State, a book published in 2008, economist Anne Morrison Piehl and sociologist Bert Useem marshal near-definitive evidence that since 2000, increased incarceration has had rapidly diminishing crime-reduction returns.