April 2008 Quarterly Forecast
> The Building Slump in Context
The Building Slump in Context
By Ryan Ratcliff
After three years of non-stop media coverage of the stunning turnaround in California’s housing markets, it’s easy to get lulled into thinking that this is the worst housing disaster since Noah and the Flood. And in some ways it is. But in terms of the decline in building activity, our current situation isn’t that different from the 1990s. Given the emerging puzzle of the strength of the East Bay’s Construction industry in the face of a major decline in new building, taking a look back at previous building slumps can give us some insight into what to expect in 2008.
For some broad context, let’s start with California. Our goal is to compare the current decline in building permits and construction employment to what happened in the 1990s, so we’ll need to use some sort of index to put the two periods on equal footing. Figure 7 presents a graph in which the cyclical peak of a smoothed series for building permits is set equal to time zero, and the level of building permits at the peak of each cycle is set equal to 100 (these peaks occurred in Feb. 1990 and Oct. 2005). This allows us to compare both overall percentage decline in building permits (100 minus the current level), as well as the speed with which this decline occurred. In late 2007, we can see that California residential building permits declined by about 60% over two years, which is about exactly the same decline the we saw in California from 1990 to 1992. However, the path we’ve taken to get to this 60% decline is much different than before. In the 1990s, the majority of this decline occurred in a precipitous drop during the first year, while our current slump has been more a slow but steady decline.
Figure 7 also includes a similar graph for the level of Construction employment, where the level of Construction employment at the peak of building permits is also set equal to 100. We see a similar pattern here: two years after the peak of building permits in 1990, California construction employment had fallen by just shy of 25%, while Construction employment has only fallen 7% over a similar period today. The big difference between today and the 1990s is of course the recession, which officially began in July of 1990 and ended in March 1991. As we’ve discussed time and time again in our national forecast, today’s construction slump is a historical oddity: a recession would in most cases have already come and gone by this point in previous construction cycles. So while we have a recession-sized decline in building activity, job losses in Construction have been moderate by historical standards because the rest of the economy has held up better this time around – so far.
Figure 7: Index of CA Residential Building Permits (Smoothed) and CA Construction Employment (SA)
Figure 8: Index of Oakland MD Residential Building Permits (Smoothed) and Oakland MD Construction Employment (SA)
With this overall story for California in mind, let’s apply the same analysis to the East Bay specifically. This involves a little bit of a fudge with the 1990s peak: the data for the East Bay only begins in January 1990, and is declining from the start. The peak of building permits may have occurred earlier, but we’ll assume it begins in January 1990, and take comfort from the California data that this assumption probably isn’t too far off. The first and most striking feature of Figure 8 is that decline in building permit activity is much smaller in the East Bay than in California in both slumps: 40-45% compared to 60%. We see the same differences in timing: the 1990s decline occurred over 1 year, while today’s decline has taken close to 3 years to get to the same place. And it’s no surprise that a smaller decline in building activity during the 1990s relative to the state average should lead to a smaller percentage decline in Construction employment. But Figure 8 once again underscores the central puzzle of the East Bay economy in 2008 – Construction employment is hanging tough at near the same level as at the peak of building activity. Interestingly, the unique history of the Bay Area gives another data point to consider: the late 1990s saw a mini-boom in East Bay building that was not paralleled in the rest of the state. As the tech bust slid into recession in 2001, we again see that recessions matter: a 31% decline in East Bay building permits over the three years from February 1999 to February 2002 saw an 8% decline in Construction employment. However, the employment decline didn’t begin until March of 2001 – the exact beginning of the 2001 recession.
Figure 9: YoY Growth in Res. Bldg. Permits and Construction Employment, Oakland MD
What lessons are we left with after comparing today to the 1990s? First, the current decline in building permits is hardly unprecedented, but has occurred slower than in previous cycles. Second, the employment impact of the decline in building seems to be exacerbated by recessions. Unfortunately, this second point also muddies the waters when considering today’s slump: is the resilience of East Bay Construction employment simply the result of a relatively smaller decline in construction activity coupled to the lack of a recession in 2006-7? If that’s the case, it may hold up. However, the fact we’ve seen an overall decline in California construction employment under similar circumstances leaves room for doubt…
Figure 10: CA Counties Percent Decline in Permits and Construction Employment from Peak
For more insight into whether the East Bay’s Construction sector can continue to shrug off the decline in building, we can look at the other counties in California to see if this disconnect between building and employment is unique. Figure 10 plots a scatter diagram of the percentage decline in construction employment off its peak versus the percentage decline in building permits in major California counties. Overall, we see the expected relationship that the counties with a bigger decline in permits are also the counties with bigger declines in Construction employment. However, we also see that the East Bay is not alone in seeing 70-80% declines in residential building permits with relatively minimal declines in Construction employment: Santa Barbara, Sonoma, and San Jose Counties all show even more extreme versions of the same trend.
This particular list of counties makes it unlikely that this commercial or heavy construction that is offsetting residential weakness – so what do these counties have in common? Figure 11 plots the level of Construction employment in each of these counties as index that is set equal to 100 at 2000’s average level of employment, as an effort to measure both the size of the boom and the bust in Construction employment. We immediately see a common denominator among these counties: each of them experienced fairly significant declines in Construction employment during the 2001 recession, while the state as a whole did not. This is of course consistent with our story of the mini-bust in the East Bay in 2001. And with exception of Santa Barbara, each of the Northern California counties is at or below its level of Construction employment in 2001: in these counties, the construction boom provided a return to normalcy in Construction, in contrast to the “overemployment” we have seen in other regions. This gives us a plausible reason why we might expect these smaller than average losses in Construction employment to hold up in these counties: less boom before means less bust now.
Figure 11: Index of Construction Employment, 2000 Average = 100
Figure 12: Decline in Financial Activites Employment Since Peak versus Total Increase in Fin. Act. from 2000 to Peak
Figure 12 attempts to apply this logic to the finance industry: are the counties with the most weakness in Financial Activities employment today the ones that had the biggest booms? For the most part, the answer is yes; however, we once again have a few counties that don’t line up with the trend. San Francisco, Sonoma, Santa Barbara, and Monterey Counties have experienced fairly significant declines in Financial Activities employment, yet did not see very big booms. Outside of Sonoma County, each of these counties has a 2005 location quotient for both Non-Depository Credit Intermediation and Real Estate Credit that are both below 0.5 – real estate related finance was not a big part their economy (small boom), but they too are suffering from the spillover effects. Sonoma County represents the odd duck, with a 2005 location quotient of 0.77 for Financial Activities (ie not a big local specialty), but a noticeable specialization in Real Estate Credit, with a location quotient of 1.61 in 2005. This is the worst case scenario – all the pain of the bust, but without the temporary benefits of the boom…
| 
