‘Rookverboden in onze bars en restaurants doen de sector geen schade aan’ was de conclusie die het ministerie van Volksgezondheid in New York trok toen ze met veel verve meedeelde dat er sinds het ingaan van het rookverbod in maart 1.500 banen in de horeca bij waren gekomen.
Een journalist in de New Yorkse City Journal laat zien waarom een dergelijke conclusie op zijn zachtst gezegd erg voorbarig is.
“The health department’s stats on jobs and the smoking ban don’t add up.”
I wouldn’t quite call the report a “damned lie,” but it is a case of using statistics in the service of blatant political propaganda. In its report, the health department employs unreliable preliminary data—heavily massaged and likely to be revised (several times, probably)—to jump to an extremely questionable conclusion about an exceedingly complex subject.
In its release, the department says that employment at city bars and restaurants increased by 1,500 “seasonably adjusted” jobs from March to June. Since the new anti-smoking law went into effect on March 30, the report declares “the law has not had an overall negative impact on business.”
There are so many things wrong with this “report,” it’s difficult to know where to begin, but let’s start with the actual number of restaurant jobs in New York from March through June. First, we need to know that the government’s monthly employment data for New York City are not based on an actual count of jobs, which would be too burdensome to do every month, but on a projection made from a limited survey of local employers.
What good are these monthly reports, then? Well, at the most they might indicate a direction the economy is heading, or something worth watching further as more exact data come in. But for long-term trends or studies, economists are much more likely to rely on job stats averaged over the course of an entire year for New York.
Not the health department. It used the monthly data to conclude something far more complex: the influence of a new law on an entire industry. For that you need not only accurate employment statistics but the ability to factor out other significant economic trends operating at the same time.
To get a somewhat more accurate picture of what may be happening in the business right now, it makes more sense to look at the jobs data that are not seasonally adjusted from this year and compare each month to the same month last year, which reduces the impact of the seasonal hiring. Those figures show that in every month from March to June, the city had fewer restaurant jobs this year than it did in the same month last year. By contrast, in both 2000 and 2001, two years when the city’s restaurant industry grew, in every month from March to June restaurants recorded substantially more employment than for the same month the previous year.