Creative Class

It’s more than a jobless recovery, we’ve been looking at a jobless decade or more, at least in terms of private sector jobs, according to Business Week’s chief economist, Michael Mandel.

Beneath this trend lies a broad and fundamental restructuring of the U.S., and virtually every other advanced economy – the decline of manufacturing and the rise of professional, knowledge-based, and creative work on the one hand, and lower-end service work on the other. This chart (via the New York Time’s Floyd Norris) depicts the shift.

Yesterday, we looked at the relationship between housing prices and income. Today, we turn to the relationship between housing prices and wages. Wages are a useful way to gauge regional housing prices because they only count money that is earned by doing work. Income, on the other hand, counts any and all earnings from investments, interest, dividends transfers, and other sources.

The graph below plots housing prices in 2009 against wage levels for 2008 (the most recent data available).

There is a clear, positive, linear, and significant relationship between wages and housing values – the correlation is 0.71 and the R2 0.51. Metros above the fitted line had higher housing prices than wages relative to national levels, while those beneath the line had lower than expected housing prices.

Near the top we see many of the same regions as in yesterday’s analysis. Honolulu is once again the greatest outlier, with housing prices exceeding wage levels with a differential of $384,290. Metros in California once again play a prominent role at the top of the list, including San Diego ($87,365), Los Angeles ($63,340), and San Francisco ($60,148). New York also registers a substantial differential of $76,896; Miami ($46,128) also has a considerable differential.

On the other hand, there are metros where housing prices were less than their incomes would predict based on the national trend. In Decatur, IL, for example, housing prices were $131,344 less than what its wage level could support based on the national trend. In Michigan, both Saginaw ($123,140) and Lansing ($119,334) had differentials over $100,000, as did two Ohio cities, Akron ($105,447) and Cleveland ($105,386). Atlanta ($86,079), Washington, D.C. ($65,446), and Dallas ($51,896) all had differentials of greater than $50,000, while Houston ($48,874), Chicago ($48,794), and Boston ($42,834) all had differentials of greater than $40,000. The difference was more modest in Philadelphia ($20,520). We’ve again omitted Detroit because it failed to report housing price data for 2009.

The past couple of days, we’ve looked at the relationship between past and current housing prices. We saw that there are some regions where housing prices have fallen more than what might be expected based on national trends, while prices have declined considerably less than expected in others.

Today, we shift gears looking at the relationship between housing price and incomes. The graph below compares median housing prices in 2009 to income per capita levels in 2007 (the most recent figures available).

Housing prices and incomes are closely associated with one another: The correlation coefficient is 0.68 and the R2, 0.46. Metros above the fitted line have housing prices that are higher than their incomes relative to the national trend, while those below the line have housing values that are less than what their incomes would predict relative to the national trend.

In Honolulu, for example, the differential was a whopping $371,777. Almost half of the top 10 regions are in California. In San Jose the differential is $120,134, San Diego ($106,625), Los Angeles ($103,278), and San Francisco ($59,633). The differential was also in the Pacific Northwest – Portland ($74,490) and Seattle ($60,848), as well as Salt Lake City ($77,526) and New York ($93,900).

On the other hand, there are metros where housing prices were significantly less than their incomes would predict based on the national trend. In Bridgeport, CT, for example, housing prices were $151,460 less than what its income level could support based on the national trend. In Cape Coral, FL, the figure was $110,460. This was also true in Rustbelt regions like Akron ($106,692) and Cleveland ($105,130) which had differentials greater than $100,000. There were also considerable differentials in two Texas cities, Houston ($93,586) and Dallas ($58,602). In addition to this, Atlanta ($50,166), Chicago ($30,337), Philadelphia ($18,699), and Washington, D.C. ($17,280) all had housing prices that are less than their incomes would predict based on the national trend. We’ve omitted Detroit because it failed to report housing value data for 2009.

Earlier this week, Foreign Policy released the latest edition of its Failed States Index (via Daily Dish’s Patrick Appel). It’s based on a database of 12 indicators of state cohesion and performance for 177 nations. So my colleague Charlotta Mellander decided to compare it to our Prosperity Institute economic development database which has a wide range of indicators for output, productivity, human capital innovation, life satisfaction, human development, and economic structure. The findings, while not particularly surprising, are nonetheless interesting. FP asks:

“[W]ho (or what) is to blame when things go bad—corrupt leaders, dysfunctional societies, bad neighbors, a global recession, unfortunate history, or simply geography itself? ”

The simple answer that comes from our analysis is development – or lack of it. Failed states not only fail on state cohesion and performance, they also fail on measures of economic development from output or GNP per capita and total factor productivity to human capital, life satisfaction, and more. And failed states apparently lag badly on the transition to knowledge-driven, creative economies.

The U.S. economy has shed 7.2 million jobs since the onset of the recession. But the economic pain of unemployment has not been spread equally, according to a new analysis by my colleagues at the Martin Prosperity Institute.

The graph below, compiled by Ulrich Atz, tracks the unemployment rate for three broad groups or classes of employment – the working class, the service class, and the creative class from 1971 to May 2009.

The report finds that:

Unemployment for all three groups has spiked since the onset of the recession.  But the downturn has hit hardest on working class. . . The working class has been hard hit by every downturn since 1971. Working class unemployment spiked from 6.2 percent in 1973 to 14.5 percent in the 1975 downturn.  It spiked again from 7.7 percent in 1979 to 16.8 percent in 1983.  It reached 12.0 percent in 1992.

In contrast, the unemployment rate for the creative class has hardly ever reached the 4 percent mark.  Unemployment rates among the working and service class are typically about 3-4 and 2-3 times respectively the rate of those in the creative class.

A closer look at monthly data (available starting in 2000) reveals that unemployment rates among the working and service classes typically move together while creative class unemployment lags the other two by several months.

The full analysis is here.

Yesterday, I compared 2009 housing prices to their 2006 baseline. Today, I turn to the change in housing prices. The graph below plots the percentage units change in housing prices between 2006 and 2009 against the 2006 baseline price.

There is a significant relationship between the two. The slope is steep, with a correlation of  -0.42 and the R2 of 0.19. Metros above the line have seen drops which are less than would be expected based on national trends, while those below the line have seen drops in excess of the national trend. The numbers in parentheses are the percentage difference between the actual and predicted values.

Under-performers: These are regions where the decline in housing prices has been greater than predicted based on the national trend. The biggest losers are metros in the Sunbelt and Rustbelt. In Cape Coral-Fort Myers, FL, for example, housing prices have declined 47.3 percent more than expected based on the national trend. For Akron, OH, the figure is 44.9 percent; Lansing, MI (-39.6 percent); Cleveland, OH (-35.4 percent); Grand Rapids, MI (-33.9 percent); Phoenix, AZ (-31.7 percent); Sarasota, FL (-29.7 percent); Riverside, CA (-29.3 percent); Toledo, OH (-29.3 percent); Palm Bay-Melbourne, FL (-29.1 percent); Sacramento, CA (-28.8 percent); Canton, OH (-28.3 percent); and Las Vegas, NV (-28.2 percent). Miami (-18.56 percent), Atlanta (-18.05 percent), Chicago (-11.72 percent), Los Angeles (-10.07 percent), and Washington, D.C. also performed worse than expected.

Over-performers: There were again a series of regions that performed better than the national trend. These are places where housing prices have held up better than expected based on the national pattern. In Honolulu, HI, for example, housing prices remain 31.1 percent above what could be expected based on the national pattern. Cumberland, MD, a suburb of Washington, D.C., has held up 30.4 percent better than expected. In Salt Lake City, UT, the figure is 29.8 percent; Bismarck, ND (26.2 percent); Beaumont-Port Arthur, TX (25.9percent); Farmington, NM (25.7 percent); Binghamton, NY (24.2 percent); Columbia, MO (22.4 percent); Raleigh, NC (21.3 percent); and Austin, TX (19.7 percent). New York (11.3 percent), Philadelphia (7.4 percent), Houston (6.37 percent), and Dallas (+4.2 percent) also performed better than expected.

Housing prices continue to fall nationally but the economic impacts of the crisis are being felt unevenly across the country. Housing values are off roughly a third from their peak in mid-2006, according to the Case-Shiller Home Price Index. Phoenix and Las Vegas have taken the biggest hits, suffering declines of more than 50 percent in the past year. Miami, San Diego, L.A., and Tampa have also been hard hit. Detroit has seen housing prices sink to mid-90s levels. Housing prices have declined less significantly in greater D.C., Chicago, Seattle, Atlanta, New York, Portland, Boston, Denver, Dallas, and Charlotte. But the Case-Shiller data only covers 20 large metro regions.

This week, I take a look at how housing prices have fared across the full set of more than 300 American metropolitan areas. The posts are based on statistical analysis by my colleague Charlotta Mellander. Today and tomorrow, I’ll look at how housing prices have fared since their 2006 peak. Later in the week, I’ll look at the relationship between housing prices and incomes and wages.

The graph below compares housing prices in 2009 to their 2006 baseline price. It’s based on “residual analysis,” comparing the change in housing prices between 2006 and 2009.

Clearly, the two are related – the correlation is 0.903 and the R2 is 0.815. But the slope of the fitted line suggests that, on average, housing values in these regions have dropped by approximately 15 percent. Metros above the line have lost less value than their 2006 worth would predict, while those below the line have lost more.

Under-performers: These are regions where housing values have slipped even more than predicted. Among large metros, the under-performers include: Los Angeles (where values are off $79,789 more than expected based on the national trend), San Francisco (-$79,029), Las Vegas ($-72,421), Phoenix (-$69,897), and Miami (-$53,021). Cape Coral, FL saw the biggest relative decline (- $111,797), followed by Riverside, CA (-$103,683), Sacramento, CA (-$91,640), and Sarasota, FL (-$82,353). Akron, OH (-$59,635) and Lansing, MI (-$57,574) also saw significant declines. Housing values were down slightly more than would have been expected in Atlanta (-$27,413), Chicago (-$16,580), and greater D.C. (-$14,411). 2009 data for Detroit were not available.

Over-performers: The analysis turned up a number of over-performing regions. By that I mean regions with housing values performed better than expected relative to the national trend. Over-performers include: Honolulu (where housing values remain $160,414 more than expected), Boulder ($72,172), Salt Lake City ($68,935), Seattle ($61,997), New York ($58,407), Raleigh, NC ($57,552), Portland, OR ($42,173), Baltimore ($39,896), Austin ($38,181), Philadelphia ($29,011), Boston ($13,644), Houston ($8,693), and Dallas ($5,661).

Stay tuned for more tomorrow.

The impacts of the economic crisis continue to be felt unevenly across the country. I’ve previously looked at the factors associated with higher rates of regional unemployment. But which places have seen the biggest jumps in unemployment since the crisis hit?

To get at this, my colleague Charlotta Mellander conducted a straightforward statistical exercise called a “residual analysis.” It’s a simple way to track how a location performs relative to the performance of all other locations. Basically, the analysis examines to what extent the initial unemployment rate in May 2008 seems to have had an impact on the change in unemployment over the last year. Technically speaking,  Mellander ran a regression analysis predicting change in unemployment over this last year (May 2008 to May 2009) as a function of the initial level of unemployment at the beginning of the period (May 2008). She then compared the predicted values to the actual values.

The first graph shows the pattern for U.S. states.

The hardest hit states are ones that were doing badly even before the crisis hit. The fitted line is steep; the correlation between the two is 0.59 and significant; and the R2, 0.345. States below the line experienced a smaller than predicted increase in unemployment levels, while those above the line saw a larger than predicted increase.

Michigan has the highest unemployment rate, but Oregon (+3.0) has taken the biggest relative hit. Alabama (+1.8), Indiana (+1.6), South Carolina (+1.6), and Wisconsin (+1.4) have also taken bigger than expected hits. North Dakota has the lowest rate of unemployment but Alaska (-2.8), Mississippi (-2.1), Arkansas (-1.2), Connecticut (-1.2), Iowa (-1.1), and Nebraska (-1.2) have done better than expected.

The second graph repeats the analysis for U.S. metropolitan regions. It excludes two extreme outliers in California – Yuma and El Centro – which started the period with 20 percent plus rates of unemployment.

The hardest hit metros are also those that were doing badly before the crisis. The fitted line is again steep; the correlation coefficient is high, 0.59; and the R2, 0.351.

The crisis has hit hardest at smaller Rustbelt metros, especially those in Indiana: Elkhart-Goshen, IN; (+7.3); Kokomo, IN (+7.2); Decatur, GA (+3.2); Sheboygan, WI (+2.7); Fort Wayne, IN  (+2.3); and Youngstown, OH (+2.2).

While Detroit has faced staggering unemployment, the difference between its actual and predicted unemployment is +1.6. Among large metros, Portland (+3.1), Charlotte (+2.2), and, San Jose (+1.9) experienced even bigger than expected increases in unemployment. Las Vegas (1.5), Boise (1.29), and Orlando (+1.29) have also been hard hit. San Francisco (+.93), Miami (+.49), L.A., Chicago (+.31), Atlanta, and San Diego (+.21) also performed worse than their May 2008 unemployment levels predicted.

Several Oregon metros took worse than expected hits: Bend-(+4.6), Eugene-Springfield (+3.8), Portland (+3.0), Salem (+2.5), Medford (+2.4), Corvallis(+1.9). Metros that border Oregon like Spokane, Washington (+0.8) and Boise, Idaho (+1.3) also have high differentials.

Three Texas cities – Dallas (-1.0), Houston (-0.9), and Austin (-1.0) – performed considerably better than expected. Minneapolis-St. Paul (-0.4) did too. Cities along the Bos-Wash mega-region – Boston (-0.4), D.C. (-0.3), New York (-0.1), and even Philadelphia (-0.3) – also did better than predicted. Surprisingly, Phoenix also outperformed expectations (-.2), albeit modestly.

College towns number among the best performers, doing much better than predicted: Champaign-Urbana, Illinois, home to University of Illinois (-2.2); Iowa City, University of Iowa (-1.81); Manhattan Kansas, Kansas State University (-1.82); College Station, Texas, Texas A&M (-1.74); New Haven, Connecticut, Yale University (-1.54); State College, Pennsylvania, Penn State University (-1.47); Boulder, Colorado, University of Colorado (-.93); Austin, Texas, University of Texas (-1.0); Ann Arbor, Michigan, University of Michigan (-.94); and Ithaca, New York, Cornell University (-.97), among others.

This past week, I’ve look at the trends in U.S. innovation, commenting on Michael Mandel’s powerful and compelling thesis about the deceleration and interruption of American innovation. With the help of my MPI team, I’ve tracked patent data since 1980, examined patent trends for U.S. resident and foreign, non-resident inventors, and looked at the geographic distribution of patenting.

Overall, the trend in patenting is up – both in absolute numbers and controlling for population. Innovation has increased over the past decade, but not at the breakneck pace of the 1980s and 1990s. There have been two dips in patenting over the past decade – the first in the wake of the tech crisis of 2001 and the second, more recently, concurrent with the onset of the housing and financial bubbles and the subsequent economic crisis.

American innovation has shifted and become more geographically concentrated. Places like Silicon Valley and Seattle have seen a steady increase in innovation while older, industrial centers like Pittsburgh and Detroit have declined significantly. Innovation in large cities like New York and Chicago has stagnated. And American innovation has grown increasingly dependent on non-resident, foreign inventors.

Today, I focus on a broader historical question: How do economic crises affect American innovation? Does innovation slow down or speed up during periods of crisis?

Joseph Schumpeter long ago argued that crises were seedbeds of innovation and entrepreneurship. Innovations developed during crises generate the gales of creative destruction that launch new technologies, remake existing industries, and give birth to entirely new ones – setting in motion new rounds of economic growth. Economists Gerhard Mensch and Christopher Freeman have examined the historical timing of innovations, with Freeman famously arguing that the pace of innovation is actually relatively constant: Innovations bunch up during crises, only to be unleashed as economic conditions are restored.

The graph above is reproduced by economist Alfred Kleinknecht. It shows patent activity from 1750 to 1970. It tracks actual patents granted from 1901 to 2005. There are clear spikes in innovative activity during the Long Depression of the 1870s and 1880s and the Great Depression of the 1930s.

The historical literature also suggests that crises are periods of significant innovation. Joel Mokyr and Naomi Lamoreaux have documented the rise of important innovations like the incandescent light, the steam turbine, and the transformer during the Long Depression. Economic historian Alexander Field finds the 1930s to be the “most technologically progressive” decade of the 20th century.

The chart below, compiled by the MPI’s Patrick Adler based upon a reading of the historical literature, identifies some of the major innovations of the Long Depression and the Great Depression. If the past is any guide, we should expect some acceleration of innovation – and particularly of the dramatic innovation Mandel wants to see – in the coming decade.

The graph below, compiled by my colleague Charlotta Mellander, updates the story, charting patents granted per 10,000 people from the 1890s to 2007. The rate of innovation rose significantly after the Long Depression. It then dipped during the Great Depression before trailing off considerably during the World War II period. American innovation rebounded remarkably in the post-war period before trailing off in the 1970s. Since the early 1980s, however, American innovation has surged to record highs. There have been two dips in innovation in the 2000s. But as of 2006 or 2007, innovation has fallen only slightly from its record pace.

So what’s happened to U.S. innovation? Like virtually every other facet of the economy, it has been – and continues to be – reshaped by globalization. As we saw on Thursday, foreign non-resident inventors have become a key element growing U.S. patenting and a big piece of the American innovation system. Beginning around 1980, non-resident inventors essentially closed the gap with U.S. inventors. By the late 1990s, they had pulled even and were at times outpacing U.S. inventors. This is part and parcel of the globalization of the economy and the fact that the U.S. is the biggest market and most innovative nation on the planet.

This has altered the American system of innovation in a deep and fundamental way – changing it from a system that for the better part of a century was based on producing and commercializing innovations to one that is more attuned to attracting inventors and innovation globally. This shift is also reflected in the changing geography and regional concentration of U.S. innovation – the decline of old, integrated, regional innovations systems in locations like Pittsburgh and Detroit and the rise of new, globally focused clusters like Silicon Valley.

Innovation is no longer an American game – or, for that matter, a game of any one nation. The countries of the world are now all part of a much more global innovation system. Strategically, this shift means from organizing to generate new breakthrough innovations to organizing to absorb innovations coming from many different sources worldwide.

The U.S. is uniquely positioned because of its size, scale, universities, and venture capital system; its sophisticated end-users and customers; and its ability to attract global talent – to harness and reap the benefits of this global system. Its major innovation clusters reinforce this advantage and they will be hard to displace. That said, for the first time, the overall rate of American innovation has come to depend on foreign inventors. Anything that might slow the immigration or inflow of foreign inventors – or redirect their inventions and patents – would undoubtedly damage the rate of American innovation.

The key question for the future is less about the slowdown in innovation and more about which people and places will prosper in this new age of accelerating global innovation.

The past couple of days, I’ve looked at the trends in overall patents and nationality of inventor. Today I turn to the regional distribution of innovation across U.S. regions.

It’s well-known that high-tech industries are concentrated and clustered in areas like Silicon Valley, Greater Boston, Seattle, Austin, and North Carolina’s Research Triangle. Paul Krugman won a Nobel Prize for his pioneering work on the relationships between urbanization, trade, and economies of scale. And Michael Porter has shown how and why innovative firms cluster.

The graph below, compiled by Scott Pennington of the Martin Prosperity Institute, shows patent trends from 1976 to 2007 for the top 10 U.S. regions. The graph identifies a clear shift in the geography of patenting. The level of innovation has fallen off considerably in older industrial regions like Pittsburgh and Detroit. It has also fallen off in Sunbelt regions like Dallas with a large presence in computers and communications and Houston with its strong concentration of resource and energy industries. On the other hand, innovation has increased substantially in high-tech regions like Silicon Valley, San Francisco, and Seattle and also in Los Angeles. Two other large regions – New York and Chicago – more or less conform to Mandel’s thesis: Both saw dramatic growth in the late 1990s followed by precipitous drops in the 2000s which erased those gains. Overall, American innovation has become more geographically concentrated and spikier.

The decline of industrial regions as centers of invention reinforces the point made by Henry Ergas two decades ago: The U.S. innovation system is skewed heavily toward “shifting” (the creation of new breakthrough technologies and products) and away from “deepening” (the application of new inventions and technologies to the continuous, incremental upgrading of older industries). The decline of GM and Chrysler – and in particular the latter’s acquisition by Fiat to gain access to new technology – stand as testimony to that. The decline of innovation and commercialization in older industrial regions means that in certain key areas of technology, the U.S. has essentially ceded the potential to develop new industrial goods and consumer products to other countries – from established competitors Germany and Japan to emerging ones like India and China – which possess the industrial infrastructures to embed them in commercial products.