My Scariest Graph

Everyone has their own candidate for ‘scariest graph of recent economic news.’ Here’s mine.

It comes from the excellent recent McKinsey Global Institute report “An economy that works: Job creation and America’s future.” The MGI team had the clever idea to plot how quickly jobs came back as the economy rebounded after each post-war recession. To be more precise, for each recession-and-recovery cycle they graphed how many months it took for the total # of jobs to get back to its pre-recession peak after GDP had come back to its pre-recession peak. There’s always a lag – jobs come back more slowly than GPD does —  but MGI wanted to see if that lag had been growing or shrinking over the years.

It’s been growing. A lot. Here’s the graph:

The cycles from 1948 to 1981 are remarkably consistent — it takes about half a year for employment to come back. And then things started to change; the employment recovery following the 1990 recession took almost twice as long as any previous one, and the 2001 time lag was over twice as long again. Given the depth and severity of the Great Recession and the achingly slow pace of job creation since its end, everyone expects the 2008 column on this graph to be by far the tallest one. In fact, there’s a lot of concern that we might not even get back to pre-recession employment levels before the next recession hits.

I don’t think it’s any coincidence that the lags get longer and longer as we head ever-deeper into the digital era. Recessions force companies to take a hard look at themselves and learn what they can do without. As computers get more and more powerful and capable businesses find that they can do without as many people. They then find that even when growth resumes they don’t need to resume hiring at anything like the pre-recession pace, since each worker they have is so much more productive and capable thanks to technology. So I don’t know how long the column for the 2008 recession is going to be on the above graphs, but something tells me it’s going to loom over the other ones like a skyscraper over a residential neighborhood.

What’s your scariest graph? And which ones, if any, contain good news? Leave a comment, please, and let us know.

  • Peter Bessman

    Oh Disqus. I wrote, perhaps, the greatest comment of all time in the entire history of THE INTERNET. It was magnificent. The Mona Lisa — nay, The Sistine Chapel of blog comments. Future historians would occupy endowed chairs devoted to its study. Then Disqus ate the comment. What a pity. Herewith a second attempt, though bereft of all divine radiance:

    If you track the time for jobless recovery since 1981, it’s been increasing by a factor of about 2.5. I don’t want to make too much of blind extraplation, but that implies 8 years for our current recovery. We’re 5 years in now, and another 3 doesn’t seem hard to swallow. And the threat of another recession before recovery is all to real.

    Meanwhile, the world is facing a looming increase in the cost of energy. There is of course the peak oil theory, though the validity of that doesn’t appear to be a slam dunk. But what is assured is that an increasing global population with ever greater levels of industrialization leads to an increase in energy demands, and if supply doesn’t increase then price will.

    Morgan Downey has noticed a relationship between the price of oil and economic recession ( I wonder how this relates to automation. Roughly speaking, humans are powered by food, while machines are powered by oil (and the other usual suspects like gas, coal, nuclear, etc). Changes in the relative costs of human power vs. machine power therefore impact the viability of human labor.

    We can broadly consider automation to be the natural result of increasing machine sophistication. Machine sophistication and machine power (fuel) on the one hand, human sophistication and human power (food) on the other… this is widly inarticulate and speculative stuff, but I can’t help but think that there is something of extraordinarily great importance at the intersection of automation, energy, and economy.

    In the ninth strategy insight from Tullet Prebon (, Dr. Tim Morgan made the claim that the real foundation of the economy is energy surplus. In this view, argicultural surplus was the original energy surplus, liberating some humans from substience farming to devote the energy to other matters, leading to specialization and exchange markets. The modern economy is additionally reliant upon surplus energy from the aforementioned usual suspects, but more attention is paid the tokenization of the surplus: money. After a lot of consideration I’ve developed a great deal of sympathy for this view. And while I’m not sure what these disparate elements add up to, I can’t shake the feeling that it matters a great deal more than we collectively expect.