Urban Job Markets Under Reagan, Bush I, Clinton, and Bush II
In a nice bit of work, G. Scott Thomas of American City Business Journals (a non-partisan straight news entity as far as I can tell) crunched the last 25 years’ job numbers; his results put last weeks’s good jobs news into context:
• Nearly two-thirds of [the 100 largest urban] areas — 63 of 100 — had fewer jobs in 2003 than in 2000, the final year of Bill Clinton’s administration. The collective loss in those 63 markets was roughly 2.1 million jobs, which was larger than the total national decline.
• Seventy-nine of the top 100 metros posted slower job-growth rates during Bush’s first three years than under any of his three predecessors: Clinton, George H.W. Bush or Ronald Reagan.
• Ninety-nine of the 100 largest markets had worse employment records under George W. Bush than Clinton. The sole exception was Honolulu.
There’s a lot more in the full story, including job numbers for each of the last four presidents. There’s even good news for conservatives. Well, at least for fans of Ronald Reagan; neither Bush fares well on the jobs metric:
Reagan emerged as the leader, with 53 of the 100 markets posting their highest job-growth rates during his administration. Thirty-eight markets enjoyed their strongest growth under Clinton, and nine reached their peak under George H.W. Bush. No markets did best under the current administration.
The flip side was dominated by George W. Bush, with 79 [of 100] markets registering their lowest job-growth rates during his tenure. Eighteen hit bottom under his father, two did worst under Reagan and one reached its nadir during Clinton’s administration.
Of course, part of Reagan and Clinton’s strong performances on these scores is attributable to the low baselines each inherited due to the 1980 and 1991 recessions, respectively. Notwithstanding that, there appears to more than that at work here (for example, Bush II hasn’t just had slow job growth, he’s had negative job growth, a feat last achieved by Herbert Hoover.) And besides, while correlation is not necessarily causation, it does make for good bumper stickers.