I’ll tell ya – I was in Philadelphia last Friday night and was amazed by how lively and genuinely urban the Rittenhouse Square are is (especially in contrast to still-determinedly un-urban Washington, D.C.). But as it turns out, if I really wanted excitement, I would have gone to one of the Philly area’s factories – because according to the Philadelphia Federal Reserve’s new monthly survey of regional manufacturing activity, the area’s industry is having one of the biggest blowout months of all time.
Or is it?
The Philly Fed survey – one of several monthly manufacturing soundings conducted and issued by regional Federal Reserve banks around the country – just reported its best headline reading in more than 20 years. The +40.8 result for November, which purports to reflect a combination of all the different performance data measured by the Philadelphia and other regional Feds for their districts, was twice as strong as October’s +20.7, and it hasn’t been in this territory since December, 1993.
Many of the main internals – those individual performance measurements – jumped to a similar extent. Indeed, the readings for new orders, shipments, and employee numbers all doubled, too. And whereas the October survey recorded a decline in the average employees’ workweek, November reported a modest expansion.
But I’ve previously written on how surveys like these – along with comparable reports issued by private sector organizations like the Institute for Supply Management – suffer from survivorship bias. Whether their scope is regional or national, they measure the performance of manufacturing companies existing today, rather than the performance of the sector as a whole (as is the case, for example, with the Federal Reserve’s industrial production index).
Nor do they report actual performance levels and their changes – like production whether measured in volume or value terms, or actual numbers of workers hired or laid off. Such real output, for example, is what that Fed industrial production index tries to gauge.
Rather, these surveys measure how many firms tell analysts that they’ve increased production or added to staff (again, for example), versus how many say they have reduced output or downsized their workforce, versus how many are reporting no change on these or other fronts.
So it will be interesting, to say the least, how these Philly Fed results stack up with other November regional Fed results, and with the private sector surveys. (We already know they differ significantly from the New York Fed’s November readings. Nor do the Philly Fed results track well with the preliminary nation-wide figures for November reported this morning by the private research firm Markit.com.)
Most interesting of all, how will all these survivorship bias-vulnerable surveys compare with the Federal Reserve’s November U.S. manufacturing production data, which will be out in the middle of next month? The Fed’s October data just came out a few days ago, and as my summary showed, revealed weakness that hasn’t been seen in more than three years.
It’s entirely possible that the new Philly Fed report is a sign of (great) things to come for domestic manufacturing, and that its off-the-charts numbers will be corroborated by other data. But it’s also entirely possible that it’s a whopping outlier – even by the dubious standards of other surveys of what’s left of American manufacturing.