Big Shift, or Shifty Statistics?

Bad metrics continue to be worse than no metrics because, as Mark Twain famously said, "It ain't what you don't know that gets you into trouble. It's what you do know that ain't so."

Which brings us to Deloitte, its Center for the Edge's Shift Index, and the "Big Shift" it concludes is happening to our economy.

I quote: "The Shift Index highlights a core performance challenge that has been playing out for decades: return on assets (ROA) for U.S. companies that has steadily fallen to almost one quarter of 1965 levels ..."

It's a shocking statistic, strongly suggesting that economic collapse is imminent (oh, wait ...), even though, as the report continues, "... while labor productivity has continued to improve," which hints at some redeeming virtues.

(Note: The full report runs 142 pages and has far more virtues and faults than I can do justice to here. It gives the ROA trend great prominence as a symptom, which is why I'm focusing on it here.)

You'll recall that the KJR Manifesto specifies consistency as one of the six characteristics of a good metric. Consistency means the metric must always go one way when the item being measured improves and the other way when it gets worse. ROA fails this test. Here's why:

On the surface, this 45-year private-sector-wide decline seems to reflect an across-the-board failure of management to do its job. It's a tempting perspective, as it satisfies our shared need to find a group of people who aren't "we" to blame for whatever we're unhappy about.

Too bad it doesn't stand up to close scrutiny.

ROA is a dubious measure, even for assessing the performance of individual companies. It's too easy to manipulate, and fails the consistency test.

But that doesn't matter. Big Shifts are a macroeconomic matter, so the question is whether ROA, aggregated across the whole economy, is a useful way of looking at things.

It isn't.

Something investors know well is that different industries have radically different asset requirements. Comparing ROA across industries doesn't work.

And yet, in 1965 the U.S. economy depended heavily on manufacturing. Since that time, as you might have heard, we decided manufacturing belongs in China. Our economy now relies much more on finance, service, and entertainment.

Interestingly enough, finance, service, and entertainment seem to have far lower ROAs than manufacturing. Might this be the Big Shift that has caused the huge fall in economy-wide ROA and not a colossal failure of management?

Nor should we conclude that building an economy on low-ROA industries is a bad idea, because why would we? GDP grew from $719 billion to $14.5 trillion over the same period of time after all, and GDP growth also has some value as a measure of aggregate economic health.

Here's what's unfortunate: I strongly support one of Deloitte's conclusions, namely, that "... the gap between potential and realized performance is steadily widening as productivity grows at a rate far slower than the underlying performance increases of the digital infrastructure," (although it isn't just a nitpick to complain that their assessment of the "digital infrastructure" has more to do with amount than with sophistication and capabilities.)

Why might this be? Here are two likely explanations, both regular themes in this space.

The first: Businesses don't integrate IT into their functioning - either the technology itself or the organization. Instead, IT delivers software that's supposed to "meet requirements," leaving it up to its "internal customers" to figure out what to do with it. That's in contrast to IT collaborating with the rest of the business to design, plan, and implement business changes and improvements ... a more enlightened model, but one relatively few companies have embraced.

The second: Far too many companies are equipped with 21^st century tools but a 20^th century workforce. We have SharePoint. We have web conferencing. We have internal blogs, wikis, and all manner of other tools that can help employees be more effective, both individually and as they collaborate in teams.

And few companies make mastering those tools even a data point in assessing employee performance.

These two factors matter greatly, both to your company's success, and to our success as a world economic power.

It's too bad I can't cite Deloitte's analysis as supporting evidence.

* * *

Disclaimer: The folks at Deloitte are smart enough to have thought all this through, and know the subject matter better than I do. I've forwarded this column to them, and will publish their reply next week if they choose to provide one.

For this week's ManagementSpeak, click here.