Here’s a chart that to me sums up the past decade. This was supposed to be the Information Revolution…but what we mostly did was build homes.
Private fixed assets are things like machinery, computers, factories, power plants, housing–all the privately-owned productive assets of the country. From 1999 to 2009, the real net stock of private fixed assets grew by 26%, the slowest 10-year increase in the post-war period, according to data from the Bureau of Economic Analysis.
That slow growth in real private fixed assets is bad enough. What’s worse, housing accounted for the majority, 53%, of the real increase in private fixed assets. By comparison, information technology equipment and software–computers, software, and communications equipment–only accounted for 14% of the increase in productive assets. If we toss in spending by on ‘communications structures’, that gets us up to 16%.
In any case, the net real increase in housing fixed assets was more than triple the net real increase in IT fixed assets. That may help explain why we are in such dire straits now—plenty of new homes, not enough investment in IT.
That’s why I’m not terribly concerned about the slow pace of recovery in the housing market. I’d rather see money go to more productive uses.
Some caveats: This analysis does not include government assets or consumer durables.
Mandel on Innovation and Growth 
Do you the data to compare this with 1989-1999? How much out of the norm is this?
Steve
Housing’s share of private fixed asset growth in 1999-09 is higher than 89-99. I’ll post figures when I’m at home.
Not sure this is that relevant as real estate prices were driven by a speculative bubble, I bet the speculative rise in prices dwarfed physical “investment” in construction or other house remodeling. Consumer durables would actually be relevant as many people started using their house as an ATM during the bubble, using the price rise to pay for cars, TVs, etc, taking out HELOCs that they used for all kinds of activity. Also, if someone were to take out a HELOC to pay for online or college tuition, I suspect that won’t show up in your “fixed asset” stats anywhere, since it’s a soft skill. That’s a big oversight as the US is a 70% service economy, fixed assets are increasingly outdated and irrelevant. As for the dearth of IT investment, I think that was the market doing its job: would you invest in newspapers that are dying off online or a company like Facebook that just breaks even or ekes out a profit? The lack of micropayments led to everyone copying Google’s ad model, which doesn’t work outside search. You’d be a fool to invest in IT with no proven revenue model, particularly when the tech dummies are too dumb to realize this. 😉
I’m not devastated that IT assets were fairly small because they’re so efficient. At the rate they’ve been getting cheaper, refreshes and upgrades probably don’t show up as huge expenditures.
But I agree that this correlates well with the fact that there’s been no boom on the demand or supply side in anything really productive.
These statistics are supposed to account for the increase in efficiency.
I guess what I mean is that information technology continues to pay dividends without a high degree of investment, but investing a lot more in IT wouldn’t necessarily pay that much more.
Good point on housing. But here is a problem with the statistics — as Corrado-Hulten-Sichel have shown. Much of the investments in intangible assets don’t show up as real private fixed assets – but as consumption (expenses). It would be interesting to see the chart once that correct is made.
As a person living at poverty level and paying half my income to rent a room in a house with five other unrelated adults, I note that millions of Americans lack “affordable” housing (see “affordable housing crisis” or “affordable housing shortage)
From my perspective as one who lacks affordable housing and who understands the basic principle of supply and demand, I believe there is not yet enough housing to make it affordable to those who need it most.
What? All that invested in IT in the late 90s wasn’t enough? That was soo productive. Time to do it again?
Hat to burst anyone’s bubble, but most of the software that was bought sat on the shelf still in it’s shrink wrap, and has never been implemented. Reason is, companies were forced to buy ‘bundles’ including stuff they didn’t really need, and in fact caused redundancy and duplication of existing software, and second because on average it costs 2-2.5x the software costs in implementation services on top, which people weren’t prepared to spend.
Also, most software sold comes with a built-in 15% or so ‘maintenance fee’ each year. Unless companies drop the maintenance, they continue to pay it even if they don’t get upgrades. This is in fact a terrible investment by most firms, they don’t come close to getting their money back/value from it. If they let maintenance lapse, they have to pay for the next release as if it were new, or at least some premium to get back on the road. BTW, there’s no way they get 15% ‘new’ software each year as it is on that maintenance contract, its 1-5% at most. In many cases the underlying code hasn’t changed in years or even decades. And most ‘technical upgrades’ deliver little to a companies bottom line, only ‘functional upgrades’ provide a real ROI.
A much better use would have been capital equipment of the automated kind. This eliminates labor entirely, or majorly reduces it. Even when the Chinese and Indians have very low labor costs, 10% of ours, having 0% still beats 10% and you avoid all the other problems and costs of an extended global supply chain. We really need to look at our accounting policies to try and allow this to happen.
JohnB, there is a lot of waste in software, no doubt, but that’s because it has so much potential to revolutionize the efficiency of most business processes. The companies who use it right gain from that, others don’t. However, it’s silly to say even with all that waste, that there’s little ROI. Yes, the yearly pricing may not be strictly justified based on how it matches up with yearly improvement, but the up-front savings from the initial switch to the software is usually so worthwhile that that doesn’t matter. And it’s silly to say that automated equipment costs 0%, it just has a lower recurring cost to go with higher fixed costs. However, if you look at the entire costs, I bet that the Indians and Chinese are so cheap that they beat most automated labor too, for all but highly specialized labor. It’s all about doing it the cheapest way possible and with the surfeit of labor in India and China, manufacturing should be done there.
“I bet that the Indians and Chinese are so cheap that they beat most automated labor too, for all but highly specialized labor. ”
I read that that’s partially true, but changing. Even in those countries, automation is at increasing, even if it’s purposefully behind the Western curve.
Well, of course all manufacturing involves some level of automation these days, even in the third world. What we’re comparing is the costs of almost completely automated processes, like those welding robots used in US car factories, and the less automated processes used in the third world, that fill the gap with cheap labor. My point is that the less automated stuff is still more cost-effective for most processes, because the third-world labor is so cheap.