The FBI reported recently that serious crime fell by 10 percent during the first half of 1999, extending the nationwide drop in crime to 7 1/2 years. This is good news, to be sure, but the FBI's national figures tell only part of the story. In some jurisdictions, crime has dropped dramatically, while in others it has gone down only slightly, or even risen. And in some cases, stark differences in crime rates exist between jurisdictions that are nearly identical in every other way.

A perfect example is found in comparing Montgomery and Fairfax counties in the Washington suburbs. While the 1999 data are not yet available for the two counties, the 1998 data show that both are fairly safe places to live, with rates for almost all offenses below the national norm. But the FBI statistics show a striking disparity in crime rates between the two. For example, although its 1998 population was 11.5 percent larger than Montgomery County's, Fairfax had 1,083 fewer violent crimes. Even after controlling for population size, which works in Montgomery's favor, the Maryland county has a violent-crime rate 2.4 times that of its Virginia neighbor. Crime by crime, a resident of Montgomery is 1.7 times more likely to be raped, 2.2 times more likely to be robbed and 2.8 times more likely to suffer an aggravated assault.

This discrepancy has emerged only in the past 20 years. During the late 1970s, Fairfax and Montgomery had roughly similar crime rates. But crime rates have since plummeted in Fairfax, while dropping only modestly, or even rising, in Montgomery.

Montgomery's average population from 1978 to 1998 was 7 percent smaller than Fairfax's, but the Maryland county had 2,400 more rapes, 6,153 more robberies, 11,770 more aggravated assaults, 38 more murders and 38,087 more burglaries.

From a social-science perspective, this discrepancy shouldn't exist, since the two counties have almost identical socieconomic profiles. According to the Census Bureau, both have low poverty rates (5.4 percent for each jurisdiction in 1995), similar per-capita incomes ($ 41,539 in Montgomery and $ 39,951 in Fairfax in 1997) and similar unemployment rates (2.3 percent in Montgomery and 1.6 percent in Fairfax in 1998).

Likewise, the racial, ethnic and family composition of the two counties is almost identical. Fairfax and Montgomery have nearly the same proportions of Asians and Hispanics, and there is only a slightly higher proportion of African Americans on the Maryland side (15.3 percent vs. 8.3 percent in Fairfax). Both counties have also seen rapid population growth and large-scale immigration during the past 30 years, which has changed them from enclaves that were almost completely non-Hispanic white to models of ethnic diversity. Nor does one county have significantly more broken homes than the other: 83 percent of the families with children in Montgomery consist of married couples, compared to 85 percent in Fairfax.

So what accounts for the dramatically different crime rates between the two counties. Is it police policy? Probably not. There is little evidence that the gap in crime rates comes from differences either in the professionalism or the amount of resources devoted to policing. In 1998 Fairfax County Police and Montgomery County Police solved 57 percent and 58 percent of all violent crimes respectively.

Is it gun-control policy? Perhaps. Maryland has stricter gun laws than Virginia, which means that criminals have a greater likelihood of being confronted by an armed citizen in Virginia. Indeed, the gap between the two counties in crimes that involve face-to-face contact with a victim (such as rape, burglary, assault and robbery) is much greater than for offenses in which such contact is unlikely (larceny and car theft, for example).

Is it criminal justice policy? This looks like the best explanation. The only significant way Montgomery and Fairfax have differed over the past 20 years is that they have been governed by state legislatures and gubernatorial administrations with vastly different approaches to crime.

While Maryland has largely adhered to older and more lenient criminal justice policies during the past two decades, Virginia has emerged as one of the toughest-on-crime states in the nation. For example, in 1994 Virginia was the first state to abolish parole for violent felons. Maryland has no such law. More recently, Virginia enacted a truth-in-sentencing law requiring all violent prisoners to serve a minimum of 85 percent of their sentences. Maryland law requires only that imnates serve 50 percent of their sentences. Indeed, looking back over the 20-year period 1978-98, for every 100 Maryland criminals in prison, 66 walked the streets as parolees; in Virginia, only 45 did.

As for capital crimes, Virginia has been far less reluctant to employ the death penalty, executing 11 murderers during the first nine months of 1999 alone. Maryland has executed only three murderers in the past 20 years.

Following the rapid drop in crime in New York City after the institution of police reforms by Mayor Rudolph Giuliani, much attention was focused on better policing as the key to controlling crime. But the sharply different experiences of Montgomery and Fairfax counties offer compelling evidence that criminal justice reforms can also be effective in driving down crime rates. Conversely, lenient criminal justice policies can exact a substantial price, not just in inner cities but in some of the nation's most affluent suburbs.

Gareth Davis is a former policy analyst and David B. Muhlhausen a senior policy analyst at the Heritage Foundation.