You’ve heard and read the stories. They say that US manufacturing productivity has continued to rise over the past 15 years, accounting for the loss of manufacturing  jobs. They tout these immense productivity gains – most often attributed to automation – as the direct result of our industrial & intellectual superiority.

These are great stories. And they’d be so much better if they were true.

I recently spoke at IMTS on the overall conditions of US manufacturing and to share my belief that a comprehensive industrial policy is our best way to begin building our way back to prosperity. As part of this presentation, I laid out the general reasons for, and impact of, the distortion of US productivity numbers. Here’s what I had to say:

The fact is that the productivity numbers reported by the US Bureau of Labor Statistics (BLS) count production performed overseas that is ‘owned’ by US-based companies as US production. And at the lower wages AND with the increased profits that those lower wages have created. This mis-measurement is known as ‘import price bias.’

If the BLS productivity numbers were all we had to worry about, that’d be one thing. But many organizations and news outlets use those import price biased figures as a basis for other important reports or materials. For example, the Federal Reserve uses BLS productivity performance measurements predominantly in their monthly manufacturing regional reports. These faulty BLS numbers also factor into many trade or economic policy and projection decisions made by government entities or companies.

Garbage in, garbage out.

It’s like preparing for a battle, and having totally inaccurate data regarding adversary troop strength and location. That wrong data could get your head handed to you. And these BLS data are likely having the same impact on us as we construct a way out of our manufacturing malaise.

An excellent explanation of the full impact of these miscalculations can be found in a presentation by Michael Mandel, called ‘The Cheshire Cat Economy: Why We Are Underestimating The Impact Of Trade.’

In this presentation, Mandel presents the ‘country’ of Brickland as an analogy for the US:

But then, he introduces the fictional low-cost country ‘LowCostia’ into the supply chain equation, and what it brings (brought?) to the productivity party:

With LowCostia now in the equation, some basic changes occur:

  • GDP drops slightly, from $1,000,000 to $990,000.
  • Employment falls from 1000 to 500.
  • Measured productivity almost doubles, from $1000 per worker to $1980 per worker.

The result of measuring this production shift by ownership rather than location means that measured productivity nearly doubles – but as a result of offshoring, not automation or greater productivity through efficiency.

The cause of this perpetual charade, in a nutshell, is that the BLS continues to use formulae that were created to measure our production in a continental economy, long after we’ve entered a global manufacturing economy. Despite some band-aids applied to correct these discrepancies, they can’t begin to present a clear picture that tells us what has actually happened, nor what is really happening today.

As Mandel concludes, we need to return to ‘production economics‘ to accurately measure not just our progress (or lack thereof), but also the ‘multiplier effect’ of manufacturing jobs on the overall economy.

In all fairness, the BLS is underfunded and understaffed. Worse, even if new fomulae were developed to offset import price bias, it couldn’t possibly accurately account for the performance and shifts in foreign employment, fluctuating materials & logistics costs, labor, and governmental subsidies. And while the Federal Reserve has tried to create a formula that offsets import price bias (see page 30 of this report), it is wildly insufficient to bring clarity to this issue.

The next time you see or read a report that cites BLS productivity numbers directly or indirectly, don’t just be suspicious – be downright skeptical. Because they’re working from a faulty premise. And as a result, these faulty data mean that:

  • US manufacturing output is overestimated
  • US worker productivity is overestimated
  • Trade deficit is underestimated
  • Global growth is underestimated

To do nothing about this means we’ll continually make decisions and form opinions based on inaccurate data that could lead to even more disastrous results than we’ve already seen. These discrepancies are serious, and we must call it for what it is – inaccurate at best, deceptive at worst.

But I’m afraid we don’t want to deal with this, or we already would have. We’re happy to just accept it, and move onto the next thing. It’s easier to whistle in the dark, and just tell ourselves that everything will work out OK.

After all, it has up until now, right? Right?