OCEAN ISLE BEACH – It’s amazing, really, just how bad news reporting can get when statistics are involved.

I’m no statistician, believe me, but I do know some basics and have appropriate reference materials. I know that gasoline prices, while higher than they were last year, are hardly the highest ever (adjusting for inflation, they remain significantly below the prices I paid when first learning to drive nearly 25 years ago).

I know that North Carolina’s average teacher salaries, far from remaining below the national average after big raises during the 1990s, have long been relatively high when the state’s lower-than-average living costs are factored into the equation.

And I know that, contrary to the mythology propagated by political activists and their journalistic enablers, school spending in North Carolina varies relatively little between “rich” urban and “poor” rural counties. The vast majority of education funds are funded from state and federal taxes, not local property taxes.

Now, thanks to some interesting work at a useful website called Salary.com, I know something else: the best “buy” in North Carolina for workers seeking to earn the highest real compensation on average – salary adjusted by local costs and living standards – isn’t to be found in Charlotte or the Triangle, even though they tend to top national lists in nominal economic measures. Instead, according to the website’s 2005 “Salary Value Index,” the most attractive locations are High Point (28th out of 188 nationwide metros ranked) and Gastonia (30th).

The Piedmont Triad comes off relatively well in this comparison, actually. Greensboro (78th) and Winston-Salem (109th) are the next-highest North Carolina metros in the rankings. Charlotte (126th), Raleigh-Durham (145th), and Fayetteville (153rd) look less attractive.

The key phrase in the previous sentence is “in this comparison.” Media discussions of statistics would benefit greatly if they always included the proviso that almost every statistic one can devise to measure and communicate reality does so selectively. It is difficult, and sometimes impossible, to devise a set of statistics that doesn’t leave something important out. That’s not an argument against using statistics, it’s an argument for using them carefully and resisting the impulse to try to reduce complex subjects down to a single number.

In this case, what are left out of formal salary/cost comparisons are the diverse, subjective value judgments of individual decision-makers. Look at the city at the bottom of the Salary.com list: New York City. This is hardly surprising; people earn more than average in the Big Apple, but their living costs are much, much higher. Still, plenty of people choose to move to New York every year. Why? Because they value other things about living in the city: the restaurants, the arts and educational institutions, Central Park, sports, dynamism, the opportunities for advancement or perhaps even a little fame.

I don’t get it. There is virtually no scenario I can imagine that would lead me to live in New York City. Such diversity of preference helps to explain why statistics are so hard to use and report well – and why, even in the best of cases, they are one limited take on a complex reality.

Hood is president of the John Locke Foundation.