Ratchet up Miami-Dade’s spectacular unemployment count
Two weeks ago this space was titled “Miami-Dade’s 1.8% unemployment is spectacular – if true.” While the widely touted figure remains spectacular, it definitely isn’t true. The real figure is a realistic 2.8%.
Moreover, even that satisfyingly low December jobless level – far under the 3.5% national level of the time and Florida’s record low 3% – is due for revision.
Strip away numbers and the fact is this: Miami-Dade’s trend is just fine, with unemployment figures falling and new jobs a-building. The national picture is good, Florida’s is better and Miami-Dade’s is better still. All highly positive.
But the way we compare ourselves to others or to our own trend line is via numbers: we’re better off, but by how much?
When kids are graded we don’t ask just if they’re doing better or worse, but by how much. Did they inch up from D to C or soar from D to A? That’s why report cards were (perhaps regrettably) born.
Unlike school grades or sales receipts, however, figures on employment and unemployment are hard to interpret and always subject to later changes.
Stick with me and you’ll see that both interpretation and the effects of time are at play in the gap between the widely hyped Miami-Dade 1.8% unemployment that we questioned two weeks ago and what is today a 2.8% reality. Don’t worry: you won’t need a calculator.
Take that misleading 1.8% county unemployment. The figure was the focus of a Jan. 24 news release from Gov. Ron DeSantis. The Beacon Council, the county’s economic development partnership, hyped the number Jan. 27. Mayor Carlos Gimenez cited it on radio last week. And it still sits on the US Bureau of Labor Statistics’ web page for Florida’s economy.
But the data should come with one of those WARNING labels, like the dangers of cigarette smoking.
It turns out that while the numbers in the Bureau of Labor Statistics report exist, the bureau itself doesn’t give them credence because they aren’t reliable. They resemble a report card where Johnny took his pen and turned a D into a B.
Actually, they’re the reverse of that: what Bureau of Labor Statistics economists do is their own adjusting, taking raw estimates and partial surveys from several sources and modeling the world to shape the most realistic picture they can of what’s actually going on.
The data that showed 1.8% unemployment exist, but that’s before the bureau adjusted them for Miami-Dade’s economy really: a seasonal economy with large shifts in resident and tourist groups that alter the working world.
Other adjustments also clean up what economists call statistical noise: outlying numbers from skewed samples that don’t reflect the total picture.
The key point is, the most reliable numbers from the Labor Department are found at state levels, where samples are larger, rather than in small clusters within states – reports segment 22 Florida economic zones. Since reports are based on samples and estimates, not true totals, the smaller the area the less likely statistically a sample is to hit numbers on the head.
“Looking at the trend is probably more important” than specific numbers each month in such state segments, said Reid Kelley, a Washington economist for local area employment statistics in the Bureau of Labor Statistics who patiently helped me solve the riddle of a 1.8% jobless report that we knew was unrealistic.
Again, the more reliable numbers are statewide, where figures are more thoroughly adjusted, especially for seasonality. And Miami-Dade is one of just seven big cities that are treated like states – the others are Los Angeles, Chicago, Detroit, New York City, Cleveland and Seattle.
Those publicly available and far more accurate state-like figures show Miami’s December unemployment at 2.8%. They clear up many doubts we raised two weeks ago, the how-could-it-be questions about Miami’s economic picture.
In particular, unadjusted figures showed 17,311 fewer persons unemployed here from October to December, while adjusted figures show a more realistic unemployment drop of just 4,402 persons.
October to December unadjusted figures also show 37,624 persons inexplicably disappearing from the county labor force in two months as tourist season heated up; adjusted figures showed a much more logical gain of 3,297 workers, not a big loss.
So, using the best estimate tweaks that economists can produce, Miami-Dade is on a great trend line, from adjusted unemployment in December 2018 of 3.5% of the workforce down to 3.1% in October, 3% in November and 2.8% in December. All those are a tick or two better than statewide figures at the same dates.
But wait: it’s not over yet. More adjustments await.
A Feb. 5 release from the bureau (which still cited Miami-Dade’s unemployment fall as greatest in the US, down to 1.8%!) noted that when new unemployment figures arrive March 4 all states and the seven cities treated as states will see every one of their jobless totals since 2015 revised “to incorporate updated inputs, new population controls, and re-estimation of models.” And after that, those new figures will be chopped up by local areas, which in turn will have figures revised for the four-plus years.
Meanwhile, the states and the seven big cities will see all employment data back to 2015 adjusted as of March 16.
Among other changes, Mr. Kelley said, economists will update population controls at the state and statistical areas within the state in a census population estimate program. That, he said, is done annually.
They’ll also revise data for state and metropolitan level employment statistics. That update could affect employment and labor force levels and employment rates.
At the same time, Mr. Kelley said, individual states could revise old employment insurance claims data, which might affect unemployment totals.
If state totals change, he explained, the bureau must massage totals for each area within the state so that together they equal the new state level.
If all this looks like economists are aiming at a moving target, note this: on Friday, the Labor Department released data showing 514,000 fewer jobs were held nationally between April 2018 and March 2019 than was previously announced, the Washington Post reported. The old data had come from surveys and the new data with lower jobs totals came from reports by every business, which take far longer to compile.
The more accurate data that came later, the Post reported, made 2019 the weakest year in job growth in eight years despite a tax cut that was touted to spur more hiring.
That’s sure a far bigger change than in Miami’s jobless actual rate, 1.8% to 2.8%.
As to why data from other Florida counties aren’t seasonally adjusted like Miami’s is when most of the state has highly seasonal economies, Mr. Kelley says he doesn’t know because data are available to do it. “I’ve asked about this before, actually.”
Nor does he know why employment and unemployment totals from the other 21 economic clusters were removed for 12 days from the bureau’s Florida Economy at a Glance pages the day after they went up. He said he’d never seen those pages – the spot where a 1.8% unemployment total still sits – until we pointed him there.
As for the 1.8%, he said, “Most of that is probably statistical noise. We did the best we can to minimize that type of thing.”
And so we await March revisions.