According to the Goldman economist, estimates of cash-out refinancing published by institutions such as Freddie Mac (FRE
), the big mortgage purchaser, are incomplete. He calculates a broader measure called "mortgage equity withdrawal," based on Federal Reserve data. Hatzius shows that, during the first half of 2002, Americans' pulling cash out of their housing assets was a huge contributor to economic growth--equal to more than 2% of disposable income.
Mortgage equity withdrawal consists of mortgage borrowing minus net spending on buying or upgrading houses. The measure includes home equity loans that are used for purposes other than home remodeling; "cash-out" refinancing, in which people take out money; and capital gains from home sales that are not reinvested in the housing market.
Since the 1980s, Americans "have become much more willing to withdraw equity from their homes, particularly when house prices are rising strongly," Hatzius observes. The withdrawn equity could go toward paying down debt or investing in other assets. But Hatzius says it appears to be going mostly into spending, judging from the low personal savings rate and small inflows of money into financial assets.
Hatzius concludes that it wouldn't take much to cause a retrenchment in consumer spending. If housing prices simply stop rising and interest rates stop falling, let alone reverse course, consumer spending will take a hit. His conclusion: "Consumers are living on borrowed time." The Web may not have made the Old Economy obsolete. But it has profoundly affected labor markets, according to a study by Richard B. Freeman, a Harvard University economist.
Using government survey data from 1998 through 2001, Freeman found that workers who used the Internet on the job worked 5% longer hours than those who didn't, other things being equal. That doesn't include the extra time that people spend checking into work from home over the Internet. Freeman said there's no way to tell whether the extra hours worked are productive.
Freeman also found that people who use the Net at work get paid more. But with Internet usage widening to all types of jobs, he predicts this difference will eventually disappear. And there's no doubt that Net usage at work is picking up steam. In 2001, 41% of all workers 18 to 65 used the Net, compared with 17% in 1997.
Freeman goes on to predict that the Internet will allow more cross-border connections between job seekers and job offerers. "A person in Russia or India has the same access to job listings as a person in Detroit," says Freeman of Internet-based recruiting. At the same time, the Web lets workers commute to work virtually. So a Russian could accept a job in the U.S. without moving there.
In the future, companies may stretch the time it takes them to make a decision about hiring, Freeman says. Since they know they have access to a larger pool of labor through the Net, waiting to find the perfect match might make sense. But for European labor markets, which are far less mobile than those in the U.S., Freeman believes the way the Net tears down borders will outweigh the effect of companies waiting for the right match. That could lower unemployment in Europe.
Labor unions' recruiting and communications with members have also become much more efficient thanks to the Internet, argues Freeman. The National Writers Union, for example, gets one-third to one-half of its new members off the Web. That means unions don't have to pay expensive union reps to reach workers.
Since their Web sites are accessible to nonmembers as well, unions can broaden their influence through information dissemination. Union Web sites publish information about laws that affect workers' rights. Over the Internet, nonmembers can learn about their rights and forward the unions' agenda independently, says Freeman. With the bursting of the technology bubble, economists have questioned whether the tech spending of the late 1990s was excessive. Not by historical standards, says a new study by Merrill Lynch & Co.
From 1996 to 2001, companies increased the amount of information technology hardware and software they had in place at an inflation-adjusted annual rate of 11.2%, according to Commerce Dept. data (chart). But in the five years before that, info tech grew at only a sluggish 7.8% rate. As a result, "during the decade ending in 2001, tech capital stock grew at one of the slowest rates for any 10-year period on record," says Bruce Steinberg, chief economist at Merrill Lynch.
After the recent slowdown, Steinberg believes there is a lot of pent-up demand for new technology. For example, companies have replaced employees' PCs every three years or so, on average. That replacement cycle has now stretched to four years, and Steinberg doesn't think it can go much further without seriously affecting companies' performance. "We won't have the spending boom of the late 1990s," says Steinberg, but he is convinced that companies will soon have to increase their tech budgets or risk their competitive edge.