DATE: Tuesday, September 2, 1997            TAG: 9708310101

SECTION: LOCAL                   PAGE: B3   EDITION: FINAL 

TYPE: Public Safety 

SOURCE: BY NAOMI AOKI, STAFF WRITER 

                                            LENGTH:   71 lines

NEIGHBORHOOD CRIME FIGURES DON'T ALWAYS TELL WHOLE STORY

Statistics can be tricky business.

And readers are quick to point it out whenever we publish per-capita crime rates that differ from their perceptions of neighborhood crime.

Retailers in busy commercial districts complain that the rates make them seem high in crime. Some residents resent being lumped in with other, higher-crime areas. Still others gripe that a nearby commercial area is driving up their crime rates.

In many cases, the callers are right.

It's not that we've got the numbers wrong. And it's not that we're trying to fudge the statistics.

The problem lies in the nature of neighborhoods, census-tract or planning-district boundaries, and statistics. They don't always work well together.

In order to fairly compare one area to another, we have to calculate per-capita rates. If we didn't, a city like Virginia Beach, with more than 400,000 people, would look more dangerous than Chesapeake, Portsmouth or Suffolk, when the resort city is actually the safest in the region.

We use census-tract or planning-district boundaries to calculate ``neighborhood'' crime rates because that's how cities keep the population figures we need to crunch the numbers.

Unfortunately, when cities are sliced up into districts or tracts, neighborhoods sometimes get cut in half or lumped in with other neighborhoods or even with commercial areas.

Short of doing a major cartographical redesign and population breakdown of South Hampton Roads' cities, there's not much we can do about the problem of boundaries.

But there are still some choices.

Last week a reader called about the front-page story on declining car thefts in Norfolk and the accompanying note: ``Commercial areas tend to have higher crime rates because rates are based on an area's population and do not account for an area's traffic.''

Why, she asked, would we report the statistics this way if we know they are skewed? Isn't there a better way?

I'm not sure. But there are other ways.

Currently, we publish the number of crimes, a per-capita rate and a percent change from the previous year for all areas where crime data are available. We do that for commercial and residential areas alike and then add the aforementioned warning label for commercial areas.

Instead, we could simply eliminate from our crime reports census tracts or districts that are primarily commercial. That might be more fair, but it also has its flaws.

Readers wouldn't get information about crime in the region's most heavily trafficked areas. And that, to me, seems an important omission.

It also begs the question of which areas would be deemed primarily commercial. The Oceanfront in Virginia Beach, the Military Circle-Janaf area in Norfolk or downtown Portsmouth are obvious choices.

But what about an area like Norfolk's East Ocean View, where there are about 11,500 residents and a host of motels, convenience stores, restaurants and shops?

Or we could include the numbers of crimes for commercial areas without calculating the per-capita rate. It is certainly a more moderate solution and might be the best, given our choices.

Without a rate, there would be no way to compare commercial areas to each other or to their residential counterparts. But since those rates are skewed and the comparisons flawed, what difference does it make?

Meanwhile, readers will continue to get crime information about all areas - residential and commercial - in the region. MEMO: Let us know what you think by calling INFOLINE: 640-5555, then

press SAFE (7233). KEYWORDS: NEIGHBORHOOD CRIME STATISTICS

Send Suggestions or Comments to webmaster@scholar.lib.vt.edu