Venture Impact 2004
Venture Capital Benefits to the U.S. Economy
National Venture Capital Association
1655 Fort Myer Drive, Suite 850
Arlington, VA 22209
By Global Insight
In late 2003, the National Venture Capital Association asked Global Insight (formerly
known as DRI•WEFA) to update its analysis of the landmark 2002 study, “Measuring the
Importance of Venture Capital and Its Benefits to the U.S. Economy.” This update using
2003 data compares with the earlier study, which used 2000 data. Global Insight was also
asked to explore new research areas, including the increasing contribution of venture-
capital-supported companies in the U.S. research and development effort, what the U.S.
economy would have looked like without the recent computer-age benefits, and a deeper
look at the innovation cycle, including its volatility and venture capital’s role. We will
also explore the relation of venture capital to the U.S. stock market and business capital
investment (CAPEX), and how the United States compares with other countries in these
areas. The results of these studies are outlined in this report.
Highlights and Study Outline
The key findings of this report are outlined below as a “top ten” list.
1. Ventured firms increased their size and share in the economy over the last three
years, despite the dot-com bust and high-tech equipment-buying downturn.
2. Venture-supported company employment is up 7%, even though the three-year
economic slump produced a net loss in U.S. jobs overall.
3. The venture-supported results showed continued solid progress, even though the
funds raised, early-stage financing, venture returns, and IPOs went through
another huge boom/bust.
4. This boom/bust is another example of the creative destruction cycle, where
hundreds or thousands of firms that push the innovation envelope do not succeed,
and go out of business or merge with the winners.
5. New ventured companies have emerged as major successes, including Google,
eBay, and Jetblue. Recent medical IPOs include AtheroGenics and Onyx, with
treatments for arteriosclerosis and cancer.
6. U.S. research and development is increasingly done by small companies, many
7. Small company medical research (estimated 70% venture supported) has risen.
8. The high-tech information revolution has helped user industries like
manufacturing, airlines, and retailing be more productive.
9. The dominant U.S. position in venture capital and the widespread U.S.
technological lead has enabled an otherwise mature, wealthy economy to widen
its income and standard of living over most other advanced countries.
10. Venture capital fuels the birth of new publicly traded companies. The dynamic
stock market that follows enhances fundraising and plant and equipment
investment for all companies.
Each of these ten points and many others are discussed in detail in the following chapters.
Chapter 1 (Three Years of Contrast in 2000–03) reveals in detail the findings of the
revised and updated database of companies supported with venture capital. The database
now estimates the sales and employment contribution, which is broken down and
analyzed by industry sector, state, and region.
Chapter 2 (U.S. Research and Development: The Venture Connection) analyses U.S.
research and development data reported by the National Science Foundation, focusing on
the small companies supported by venture capital. The research and development trends
of the key venture-supported sectors of hi-tech computer and other information
technology and medical are highlighted.
Chapter 3 (The 1990s High-Tech Productivity Boom: Focus on the Users) details the
results of the Global Insight model of what the U.S. economy would have looked like
without the high-tech boom. In particular, the high-tech user sectors are modeled to show
the additional productivity, sales, and jobs brought about by their recent investment in
Chapter 4 (A Historic View of Innovation) highlights innovation from a historical
perspective, looking at examples of creative destruction in earlier technological
boom/busts and the progress that followed. Also discussed is how some modern business
successes, often adopting new technology, were born even during down-cycles for the
economy and investment.
Chapter 5 (Venture Investing: a Volatile Input to Business) traces the venture capital
links to business capital investment and a successful stock market. The benefits of a
buoyant stock market and strong capital investment, and the hi-tech contribution to this in
the United States are also discussed. Worldwide comparisons of venture capital
availability, stock market size, and capital investment are made, and their relation to
economic growth worldwide is analyzed.
2000–03: Three Years of Contrast
National Slump, Yet Venture Capital Firm Growth
The venture contribution to U.S. jobs, economic growth, and technological progress has
steadily climbed over the last three years, despite turbulent times. For the nation overall,
private employment slumped over this period. Venture-supported U.S. companies by key
measurements, including jobs created and sales, ignored the national recession and
continued their expansion. The strong growth of the heavily ventured medical sector
explains part of the growth. However, ventured companies outperformed their un-
ventured counterparts in job creation in every industry sector
The recent success and growth of ventured companies stands in contrast to venture
financing activities. These slumped after the 2000–01 boom, though in all likelihood,
only temporarily. Despite the reduced new venture funding, new seed capital is still being
planted. Companies recently venture supported have moved forward with Initial Public
Offerings (IPOs) of stock, or reached a value level where they were acquired by public
companies. Eventually, the most successful will also advance into the ranks of biggest
and most valuable firms with the most valued products. They will become the Intel’s and
Microsoft’s, The Amgen’s and Medtronic’s, the Southwest airlines and Home Depot's of
New Stars in the Venture-Supported World: IPOs in the Last Three Years
The recent IPO market is not dead. Since the prior study was released, there have been
186 venture-supported firms IPO’d. While the anticipated Google IPO will be the largest
market capitalization, it is not the largest employer, and it is not clear if it has the largest
sales. Seagate of California, founded in 1979 as a 5 ¼ disc drive maker, had sales of $6.5
billion in 2003, and employs over 10,000 within the United States. Over the past three
years, other computer hardware manufacturers have lost, on average, 14% of their
workforce; Seagate has added about 8% or nearly 1,000 workers. Today, Seagate has
expanded into disk drives ranging in size from 20 to 180 gigabytes. “Seagate's
technology advantage has enabled the company to consistently set and then shatter world
records with the highest performing disc drives in the industry. And it also provides
drives with capacities of up to 180 gigabytes capable of storing information equivalent to
a stack of typed pages three times the height of the Empire State Building.”
Seagate may have the most employees and highest sales in the recent venture capital
(VC) IPO firms, yet it is Google that is generating the most media frenzy. Google
employs 1,900 workers in three locations in California, and is aggressively expanding
around the world. Sales are unknown at this point, but are estimated at $0.5–1.0 billion.
Recent suggested valuations show that the marketplace expects exponential growth for
many years into the future, based on the type of scalable business model the firm has, and
some history. Over the past three years, employee growth has increased by a factor of 10,
and sales grew by a factor of around 40–80. With growth rates like these, no wonder Wall
Street is clamoring for an ownership share, even though Google does not need the cash
infusion. To be fair, neither Seagate nor Google have achieved all of this success on their
own. Within our database, we show two ventured firms, which have been acquired by
Google. Seagate, likewise, has acquired at least four other ventured companies. This
strategy is a time-tested successful business model. Many ventured firms eventually are
swallowed by larger entities, both ventured and un-ventured, and in some cases, making
it more difficult to count the subsequent jobs and sales contribution.
New Stars in the Venture-Supported World—Mergers in the Last Three Years
Johnson & Johnson (J&J), which acquired 35 venture companies at the time of our last
report, has since acquired many more. Of the six major acquisitions J&J lists on its
company information homepage, all six companies acquired over the last three years
were originally VC-backed firms. J&J continues to heavily rely on new idea generation
and research founded and funded at the venture capital private firm level, then as the
products become ready for market, J&J purchases the firm outright, using its size and
experience to shepherd the product through the expensive clinical trials and the marketing
and distribution of the product.
Figure 1: Ventured Companies Acquired by J&J
BabyCenter HealthCare Services Internet Information Web Portal
Inverness Medical HealthCare Products Diabetes Self Test Products
ALZA Corp. Biotech Drug delivery based pharma products
Tibotec Therapeutics Biotech HIV & AIDS research
OraParma Biotech Manufactures oral healthcare products
Scios Inc. Biotech Applies recombinant DNA technology
Past mergers still account for major portions of J&J business. Often J&J will purchase a
firm, but continue to run the business under the old name. A family of companies lists 58
U.S. domiciled firms, all 100% owned by J&J, but functioning as quasi-autonomous
subsidiary firms. Interestingly, the J&J Company Timeline Web page lists one major new
innovation over the past three years: a coronary stent. This product was developed by
Cordis, a former ventured company, which J&J purchased in 1994, but still operates it
under the Cordis name.
The Global Insight Ventured Company Database
Our latest findings show that VC-supported companies were directly responsible for 10.1
million jobs and $1.767 trillion in sales in 2003. This represents 9.4% of total U.S.
private sector employment and 9.6% of company sales. It also represents a 6.5% gain in
jobs and an 11.6% gain in sales over a three-year period. Both measures outperformed the
national numbers as the country waded through a recession. National private employment
shrank during that time to -2.3%, and U.S. company sales only rose 6.5%. In effect, for
ventured companies as a whole, there was no recession. These results reflect better small
company growth prospects and high venture penetration in some of the most
technologically advanced sectors, particularly medical, which grew rapidly during this
Figure 2: Economic Impact of Venture Capital 2000 vs. 2003
2000 2003 % change
US Jobs 9.5 10.1 6.5
US Revenues 1.6 1.8 11.6
The findings come from a database of 26,494 companies. The database enables us to
break down the ventured company performance and contribution by:
1) industry (Money Tree’s major employment industry sectors).
2) state and region.
Further details are in Appendix A
Key Findings by Industry
Employment by Industry
Job results are striking, as ventured companies added to payrolls at a faster rate than their
non-ventured counterparts in every one of ten major categories (Figure 2). Two
industries, computer hardware and services and semiconductors and electronics,
experienced net job loss for ventured companies, reflecting a slump in the recent
recession. The overall industry net job loss—including un-ventured firms—in each of
these two sectors was greater. Health-care services job growth was very similar over
2000–03, but ventured firms grew slightly faster. For the remainder of the industries,
ventured companies grew appreciably faster than the national counterpart, and in six
industries, the national counterpart shrank while ventured companies expanded.
Figure 3: National and VC Employment by Industry
30% US Employees VC Employment
20% 17% 16%
10% 5% 7%
2000-2003 Employment Growth
Not only did ventured companies grow faster than their national industry counterparts,
but the sector mix was far more favorable to ventured companies. In the sectors where
ventured companies were growing the most rapidly, compared with the national averages,
ventured companies were more represented in those industries (Figure 4)—computer
software is the best example of the sector mix advantage of the ventured companies. In
the computer software industry, ventured companies added jobs 24% faster than the
national average, growing 16.5%, while the national industry as a whole lost 7.7%.
However, the exposure to this industry is widely different in our two aggregates.
Computer software firms comprise over 10.0% of all ventured companies, yet only 0.9%
of all firms in the United States. This dramatic exposure differential is magnified in the
third column, as ventured companies employ 88% of all computer software workers.
Figure 4: Employment Shares of Totals
US Share VC Share VC / US
Biotechnology 0.5% 3.4% 54%
Business/Financial 19.7% 8.1% 3%
Communications 1.4% 8.2% 47%
Comp. Hardware & Services 1.5% 11.2% 59%
Computer Software 0.9% 10.3% 88%
Healthcare Products 0.7% 5.6% 66%
Heathcare Services 12.7% 6.5% 4%
Industrial Energy 19.8% 17.8% 7%
Retailing and Media 42.0% 23.6% 4%
Semiconductors & Electronics 0.8% 5.3% 54%
Figure 5: Share of Total Employment
8% 13% 1%
This pattern occurs many times—of the top three ventured companies that are over
weighted, compared with the national average, the employment growth differential is the
highest in two. Likewise, in the industries where employment growth is weakest, the
ventured companies are generally not as highly represented.
Figure 6 also illustrates how venture capitalists engage in and lead high-growth
industries. Ventured companies judged correctly which industries would grow more
rapidly, and even within industries, they placed their money on the fastest growing firms.
Clearly, VC financiers are efficient movers of capital. Going further, in some respects,
they helped in the creation of these fast-growth industries, bringing new ideas to market.
Figure 6 breaks down those industries that are significantly more represented in the VC
universe, and listed first and last is retailing and media, which is underrepresented. There
are ventured retail companies, including Home Depot, Office Depot, and Staples, which
have, in some cases, offered a creative business model that is copied. However, most of
retailing and virtually no media companies are ventured.
The first three industries show rapid growth in sales and employment, compared with the
rest of the United States. The next four industries have great employment gains, but not
significant sales—two of these industries typify this pattern. Both biotechnology and
health-care products are “breeder” industries—they both have significant employment
rates, as they are primarily concerned with research and development, and are then
commonly acquired by other firms once the concept is ready to market.
Figure 6: Ventured Companies, Compared With the National Averages
Share Difference Growth Differential Growth Differential
Comp. Hardware & Services 9.7% 13.1% 15%
Computer Software 9.4% 24.2% 26%
Communications 6.8% 23.2% 9%
Healthcare Products 4.9% 18.2% 3%
Semiconductors & Electronics 4.6% 16.1% 5%
Biotechnology 2.9% 17.2% 6%
Industrial Energy -2.0% 10.1% 6%
Heathcare Services -6.2% 0.5% 1%
Business/Financial -11.6% 5.0% 0%
Retailing and Media -18.4% 12.6% 12%
Sales by Industry
Ventured companies outperformed the national economy in overall sales growth. Like
employment, ventured companies increased sales more rapidly than their national
counterpart in every industry category. In one industry, VC sales fell, but national sales
fell by a greater rate. There were several industries where the national growth rates fell
while ventured companies expanded. However, sales are not the ideal unit of
measurement. Economists prefer GDP, GSP, and other value-added type measurements.
Value-added statistics measure the added input at each step in the marketing cycle.
Imagine that a ventured company makes a product from nothing and sells it for $1.0
million to a middle man/wholesaler type firm, who in turn resells it for $1.1 million to a
retailer. Both firms have nearly identical sales volume, but the wholesaler has a cost of
goods of $1 million and is only adding $100,000 in value to the product. In simple sales
terms, the firms are quite similar; in value-added terms, the ventured companies are
providing ten times the value to the economy the wholesaler firm is providing.
Once again, the favorable industry mix of the ventured companies vis-à-vis the national
averages is in the overall total sales growth. Ventured companies expanded their sales by
11.2%, whereas nationally, sales only expanded 6.5%. Yet, looking at Figure 7, this
appears to be at odds with the individual industries, as sales growth rates in the ventured
companies only appear to dramatically exceed national sales in half the industries, and
are nearly even in four industries. As illustrated in Figure 6, those industries where VC
sales outstripped national sales (column 3) are the same industries where ventured
companies tend to be found (column 1).
Figure 7: National and VC Sales by Industry
15% 11% 12% 11%
US Sales VC Sales
2000-2003 Sales Growth -16%
Wages by Industry
Wage growth by industry shows yet again how ventured companies tend to cluster in
fast-growing and higher-paid industries. Figure 8 is sorted left to right according to the
industries that ventured companies are over-represented as expressed in Figure 6. The
wages in the industries to the left of the chart are growing more rapidly than the
industries to the right of the chart, with the exception of computer hardware and services.
Once again, VC financing shows its hand, as capital tends to be lent to firms in
occupations and industries where employees are rapidly being added, and those that saw
rapid appreciation in their wages. In the following chart, the five industries with the most
rapid wage gains are communications, computer software, biotechnology, health-care
products, and semiconductors. In the VC universe, 33% of all VC employees are in these
five industries, yet in the national economy, these five industries only employ 4.2% of all
workers. More simply stated, a worker who is employed by a firm founded with VC is
more likely to see their wages rise more rapidly than the national average.
Figure 8: 2000–03 Wage Growth by Industry and National Wage Growth vs.
Ventured Company Wage Growth
18% 1 8% 2000-2003 W ag e G ro w th
11% 11% 1 1%
Latest Findings by State
Total employment growth rates over time by state showed some correlation with the VC
dollars invested in the state over the same period; however, limited to 2 data points for
2000 and 2003 for actual ventured company employment, our finding were not
compelling enough for further testing. However, several other measures were significant.
GSP growth rates over 1998–2003 showed correlation with VC funding per capita per
More interesting and perhaps more relevant to the impact of VC in the United States is
the correlation between our proxy for productivity growth and VC funding. Figure 9
suggests the more VC funding per worker in a state, the more likely that state is able to
grow output per worker over time.
GSP/Employment: Average Annual Growth 1995–2003
annual growth 95-02 and VC per
0 200 400 600 800 1000 1200 1400
The upward sloping line above supports our working hypothesis that VC-funded
companies produce more than their share of idea generation, the foundation of what
eventually shows up as aggregate productivity growth. The fit might be even stronger if
the three outliers in the upper left quadrant, DE, OR, and RI, were omitted.
There is also a strong relation between VC input and the level of output per worker.
Figure 10 uses the horizontal axis to create relative size bubbles by state. Therefore, as
VC funding per capita per state increases, states move left to right, and the bubble size
Figure 10: Higher Output per Worker in the Big VC Supported States…
100,000 NJ MA
GSP per worker
and VC per
0 200 400 600 800 1000 1200 1400
Related to the theme of higher productivity gains is the level of output per worker. Figure
10 illustrates that the states that have high rates of VC funding per worker are the same
states that have reached a higher level of output per worker. The four outlying states of
DE, NY, CT, and NJ all benefited from the financial boom of the 1990s without direct
VC input. If ventured companies contribute more than their share of new ideas, operate in
high value industries, and generate higher wages, it is no surprise that each VC workers
would contribute more GSP or value-added output per person.
The following chart shows wage levels, compared with VC input.
Figure 11: …Means Higher Wages
Average annual wages
and VC per capita
0 200 400 600 800 1000 1200 1400
It is important to create jobs. It is equally important to create well-paid jobs. So, it should
be no surprise that the states with the most productive workers are also the states with the
highest wages, as shown in Figure 11. As in Figure 10, the relation is even stronger if the
East Coast states benefiting from the 1990s’ financial boom are omitted.
In summary, the findings on VC-supported company performance are surprisingly
favorable for the last three years, given venture capital’s image of supporting high-tech
companies, many of which languished or failed during this period. Direct venture-
supported company employment rose 7% over this period, to a record 10.1 million jobs.
The total direct estimated sales of venture-supported companies is $1.8 trillion, a 12%
gain during this period.
U.S. Research and Development: The Venture Connection
U.S. research and development (R&D) is the envy of the world. Academic- and
government-sponsored research centers often contribute to the new ideas that make
ventured companies work. However, we find that ventured companies, often in
partnership with academics or founded by academics, are doing more and more of total
U.S. research. First, the increasingly ventured small-company sector is doing a rising
share of U.S. research. Second, the heavily ventured IT and medical sectors are
performing an increasing share of total U.S. R&D. Third, as some of the originally small
ventured firms grow to be among the biggest in their industry, they remain the leaders in
R&D. Of the top 50 firms in U.S. R&D, 41 were either ventured themselves or were
major acquirers of ventured firms.
Total U.S. R&D is projected to be $283.8 billion, or 2.61% of GDP, in 2003, among the
highest rates in the world. A significant amount of this is done by the government,
including military, universities, and other non-business sites. By and large, though, U.S.
R&D is performed by U.S. industry, which reached $193.7 billion in 2003, or 1.78% of
GDP. Most of the analysis that follows will discuss the portion of total U.S. R&D
performed by private industry. However, there is an additional amount of direct and
indirect support, including by ventured companies, for research performed by universities
and other non-business researchers. Most of these findings are based on reporting by the
National Science Foundation.
U.S. Research is Increasingly Done by Smaller Companies
The share of U.S. R&D done by firms with less than 500 employees has risen from 5.9%
in 1984 to an estimated 20.7% in 2003, a truly remarkable increase in share.
Figure 12: Rising Small Company Contribution to R&D
(% of Total U.S. Industry-Performed R&D 1984–2003 est)
Small Co. (Millions of $) Small Co. (% of Total)
Source: NSF, R&D in Industry: 1991–2000, Tables A3 & A4
NSF, Preliminary Release, 2001 and 2002, Tables A3 & A4
NSF, Infobrief "U.S. R&D Projected to Have Grown Marginally in 2003"
At latest count, the dollar value of small company R&D spending rose from $4.4 billion
in 1984 to an estimated $40.1 billion in 2003, about a nine-fold increase. In terms of
growth rates, this was up at a 12.3% annual rate for the entire period, far higher than the
5.1% average annual gain for U.S. R&D as a whole. The research-driven startup has been
a staple of the ventured industry, and is actually shifting the entire mix of U.S. R&D
toward smaller firms—our earlier study showed that ventured firms, adjusted for size,
spend over twice as much on R&D as unventured firms. Complete breakdowns by size of
firm and sector are limited by confidentiality requirements, but the fragmentary evidence
shows that small firms in the venture-dominated information technology and medical-
related sectors are major contributors to these trends.
In 1998–2002, within the smallest company group, companies with 5–24 employees
performed 16.0% of the total R&D spending, 25–49 performed 11.9%, 50–99 performed
20.3%, 100–249 performed 29.4%, and 250–499 performed 22.4%. Since 2000, there has
been a significant downturn in the smallest of companies, 24 or less employees, but the
small sector in total has otherwise held its share. The downturn in the very smallest R&D
effort may relate to the massive falloff of so-called “seed and early” venture funds,
comprising only 6% of total VC funds placed since 1980. As venture investing recovers
from its “seed and early” downcycle, this very small company “seed money” research for
new ideas and feasibility should recover strongly.
Big Firms Have a Venture Connection as Well
VC was in its infancy 30 years ago, and grew slowly at first. Even so, many of the
ventured companies founded in the early period of 20 years or more ago have quickly
grown from small private companies to among the largest in the country, as detailed in
earlier sections. In many cases, they have also become the largest in total dollar R&D. Of
the top-ten U.S. R&D spenders (NSF, 2001), Cisco ($3.922 billion), Intel ($3.897), and
Microsoft ($4.379), all founded in the postwar period, were directly ventured.
Of the other top ten, J&J ($3591 billion) and Pfizer ($4.847) have a massive record of
acquiring ventured firms. J&J alone was the eventual acquirer of over 30 such firms. In
fact, much of the entire drug, biochemical, and medical device sectors operate on a basic
business plan of acquiring either the small research-oriented firms themselves or their
products or patents. Motorola ($4.318 billion) is also a substantial venture acquirer.
Rise of Information Technology and Medical Research
The sector R&D trends of the heavily ventured IT and medical industries are clear: they
are performing an increasing share of total U.S. R&D. Biotech R&D spending grew from
4.28% of total funding to an estimated14.51% of total in 2003. High-tech crept up from
30.53% of total to 32.02% of total by 2003, despite the large dip from 1989–92, when it
averaged 20.83% of total R&D spending. (Note: in the National Science Foundation
R&D classification, medical equipment falls in the high-technology category. Biotech
includes all research on drugs and medicines, as well as the broad category scientific
Figure 13: Medical and Information Technology R&D Shares
Biotech (% of Total) High-Tech (% of Total)
Source: NSF and Global Insight estimates
Nominal R&D funds for Biotech companies grew at an annual rate of 12.1% in 1984–
2003. Within Biotech, R&D for scientific services grew faster, at 19.0%, while
pharmaceuticals and medicines were a bit slower, at 9.0%. The rate for high tech was
5.4%, while the total funds grew 5.1%. Within high tech, navigational, measuring,
electromedical, and control instruments was the fastest growing sector (12.4%), followed
by software (10.7%), then semiconductors and electrical components (8.3%).
Communications and computer equipment both actually received less R&D funds in
2003 than in 1984 (-1.5% and -3.9%, respectively). While the precise calculations of VC
R&D are beyond the scope of this study, the venture contribution is clearly large. Venture
high-tech activity is estimated at 65–80% of the three high-tech categories. The venture
share of medical is smaller. The large pharmaceutical companies have been around for a
long time and are generally not ventured (Many are major acquirers of ventured
companies of their research as noted elsewhere). The venture share of successful small
medical companies is estimated at 70%.
The increased penetration of small company research is most striking in the biotech
sector. The small company share of biotech research has expanded massively from 3.2%
in 1984 to 39.4% in 2003, while the share of the largest companies (25,000 or more)
shrank from 30.7% to 17.6% in 2003. We do know that of 2,842 total ventured medical
device/biotech companies in the Global Insight database, there are 2,740 small
companies, defined as less than $100 million in estimated sales.
Figure 14: Small Company Biotech Research is Booming
(Percent of Biotech Research Performed by Small Companies)
Biotech Small Co. (% of Biotech) Biotech Large Co. (% of Biotech)
Source: NSF and Global Insight estimates
Small company biotech R&D spending rose from $103 million in 1984 to an estimated
$11,1 billion in 2003, an astounding 108-fold increase. In terms of growth rates, this was
up 27.9% for the entire period, far higher than the 12.1% average annual gain for biotech
R&D as a whole. In stark contrast, large company biotech R&D spending rose at just a
8.9% clip during the same period. The increasingly common drug development model is
for the invention and new idea to be developed by a small startup firm. Later, the patent
or other rights, or the firm itself is bought by the large pharmaceutical firms who bring
their comparative advantage in final testing regulatory approval and marketing. While we
have tracked the actual acquisitions of small drug companies and accounted for their
activity, the R&D data suggest there is also a large acquisition of patent rights or other
Figure 15: R&D Share of Small versus Large Hi-Tech Companies
High-Tech Small Co. (% of High-Tech) High-Tech Large Co. (% of High-Tech)
Source: NSF and Global Insight estimates
Not to be left behind, nominal R&D funds for the smallest high-tech companies grew an
estimated annual rate of 8.5% in 1984–2003. The rate for the largest high-tech firms
(>25,000) was 2.7%, while for high-tech R&D spending in total, it was 5.4%. High-tech
small companies (<500) increased from an estimated 12.4% of the total dollars spent on
high-tech R&D in 1984 to a projected 21.6% in 2003, while the share of the largest high-
tech companies (25,000 or more) shrank from 43.6% in 1984 to 26.8% in 2003. This is
still an admirable gain for small companies. However, it is worth noting that this is a
sector where most of the big players like Intel, HP, Microsoft, and Cisco were ventured.
IBM stands out as the largest exception, though it has done major venture acquisition as
described in our prior report.
Small Company Medical Research is Booming
Yet, if small companies are increasing the growth engine for R&D, then biotech could
soon be the share leader of an ever-expanding pie. Small biotech companies only
accounted for 1.2% of the funds allocated to small firms in 1984, while small high-tech
firms garnered 64.4%. By 2003, however, the two sectors were almost even in their
respective shares of the funds given over to small firms for R&D—28.6% for biotech
firms and 33.4% for high-tech companies. As less and less R&D dollars flow to more
mature industries such as computer and communication equipment, expect biotech to
outpace the stalwart of high-tech, software, and take over as a key venture capital
Figure 16: Biotech and Hi-Tech Shares of Small Company R&D
Biotech Small Co. (% of Small Co.) High Tech Small Co. (% of Small Co.)
Source: NSF and Global Insight estimates
Information technology is a long way from dead. Recent sales data show a sustained
pickup—business investment in and consumer expenditures on high-tech equipment
growth is expected to be in the double-digit range through 2004–05, after exiting the
recent slump in 2002. The medical field showed even more strength, however, with an
acceleration of activity through the recent recession. We may be in a period where cross
fertilization and earlier work in gene mapping, etc., will permit a further acceleration of
the creation of new medicines and devices.
The 1990s High-Tech Productivity Boom: Focus on the Users
We have already documented the venture contribution to the high-tech (information
technology) industry. This chapter documents the high-tech industry’s contributions to
national productivity. We also look at detailed examples of how user industries apply
technology. Finally, we use the Global Insight large-scale model of the U.S. economy to
determine what major industries, and the U.S. economy as a whole, including jobs, would
have looked like without the high-tech investments made in the 1990s.
What is Productivity?
Productivity is the output of each worker in the U.S. economy, sometimes also referred to
as labor productivity. It is usually expressed as a growth rate. U.S. statistics count the
productivity of the nonfarm private sector, rather than every worker. They also express
the labor in hours worked rather than the number of workers.
Another less-used measure of productivity is the so-called Total Factor Productivity
(TFP). This measure estimates gains from more investment, gains from labor skills, as
well as the number working, and estimates a TFP residual not produced by these factors.
Below is the labor productivity measure.
Why is Productivity Important?
“Productivity,” Paul Krugman wrote in a famous essay 13 years ago, “isn’t everything,
but in the long run it is almost everything.” By this, he meant that when productivity
growth is high, a country’s central economic challenges—budget deficits, balance-of-
payments deficits, inflation, and income redistribution—become tractable, but when
productivity sags, these problems become intractable. Productivity is the key statistic
determining a country’s long-run vitality. It essentially measures all of the country’s
economic progress not due to the ongoing expansion of the workforce. Productivity
growth accelerated in the second half of the 1990s, after nearly 20 years of sluggish
growth. The economy underwent fundamental changes with the proliferation of
information technologies, leading many to term it the “new economy.”
Former Fed governor Laurence Meyer has gathered labor productivity estimates dating
back to 18891. Admittedly, the quality of the data is not as good for the pre-World War II
period, but it does help to put today’s innovations in information technology into
Figure 17 presents a summary of productivity growth over some interesting sub-periods
of the more than a century 1889–2002 period, judgmentally identified by Governor
Meyer. Meyer notes that average labor productivity growth over the entire period was
2%, with three periods of above-average growth and three periods of below-average
Remarks by Governor Laurence H. Meyer, before the New York Association for Business Economics and The
Downtown Economists, New York, June 6, 2001.
growth. Most of the periods span relatively long periods, with the exception of one ten-
year period (1917–27) and the current period (which has probably not run its course).
Above-average growth averaged close to 3.0%, while below-average growth was in the
1.5% range. Meyer concludes that this is a “new economy again.” That is, this is “another
period like others over the long span of American economic history, during which a
bunching of innovations has propelled the economy to a higher rate of growth for a
Figure 17: Labor Productivity Trends
(Compound Annual Growth Rate)
1889-1917 1917-1927 1927-1945 1945-1973 1973-1995 1995-2002
Post World War I
The 1917–27 productivity boom followed the first auto assembly line (Henry Ford 1913)
and the proliferation of the mass-produced, low-cost automobile. Roadways and all the
retail, movietheater, and other services of a mobile economy followed. Productivity
soared. Broadcast radio, electricity, and mass-produced household and business electric
appliances became widespread. Modern “skyscraper” construction and air travel
emerged, and Lindbergh made his famous transatlantic flight. Labor productivity growth
averaged 3.8% per year, the highest growth rate for the six segments.
Post World War II
Slow productivity growth in 1927–45 accompanied the Great Depression and World War
II era, while two factors combined to boost productivity in the years following the war.
First, output had dropped so far during the Great Depression that simply returning to
trend growth required a period of faster economic growth. Second, the economy
benefited from a wave of innovations, included the building of the interstate highway
system, antibiotics and other medical advances (penicillin came earlier), transistors, and
the calculators and mainframes that used them, and commercial jet aviation. In 1945–73,
annual labor productivity growth averaged 2.8%.
Post Oil Embargo
Productivity began to slump again in the early 1970s. Higher oil prices undoubtedly
played a role in slowing output during the 1970s. Other possible explanations include a
cooling in the rate of innovations. For information technology, there was a lull between
mainframes, calculators, etc. of the earlier era and the 1990s boom. Slower growth of
workers’ skills and greater government regulations may have contributed. In 1973–95,
labor productivity growth averaged only1.4% per year.
While there are still unanswered questions about why productivity gains slowed after
1973, a consensus has formed in academia that the surge in productivity growth we have
enjoyed in recent years (3.0% since 1996) is coming from investment in information
technologies (i.e., computers, software, and communications equipment). Dale
Jorgenson, in his “2001 Presidential Address” to the American Economic Association,
summarized this research as follows: “A consensus is building that the remarkable
behavior of information technology prices provides the key to the surge in economic
growth.” According to Jorgenson, Intel’s shift in its product cycle for semiconductors
from three to two years was a pivotal event.
Numerous studies have used growth to explain the step-up in labor productivity from
1973–95 to the post-1995 period. The conclusions of Steven Oliner and Daniel Sichel,
the Economic Report of the President, and Jorgenson, Mun Ho, and Devin Stiroh were
summarized by Baily. While all three use slightly different data to support their analyses,
there are fundamental similarities in their conclusions. As in Baily’s analysis of the
earlier time period, information technology was the largest identifiable factor
contributing to labor productivity growth.
Meyer, Baily, and others look to the bunching of productivity-enhancing innovations,
combined with a favorable U.S. economic environment to explain. In Baily’s words,
“rapid advances in computing power, software and communications capabilities formed a
set of powerful complementary innovations.” An increasingly deregulated U.S. economy
created a highly competitive environment that drove out inefficiencies, displaced low
productivity firms with high productivity ones, and forced the adoption of new
innovations in order to survive. While the new innovations were available globally, the
highly competitive environment may explain why U.S. productivity rates benefited more
from them than other world economies.
Kevin Stiroh found that the recent productivity revival is broad-based, with nearly two-
thirds of the 61 industries in his detailed industry analysis showing accelerating
productivity gains.2 This finding knocks down Robert Gordon’s objection that
productivity growth was confined to durable goods manufacturing. Stiroh also found that
productivity growth was concentrated in industries that either produced information
technologies, or used them intensively, the latter further discussed below. Thus, Stiroh’s
industry analyses supports the growth accounting conclusions that information
technology capital was a significant contributor to the post-1995 productivity surge.
User Industry Benefits
Retailing has been transformed, not only by the Internet revolution, but also by other
technology, including, at point of sale, bar code scanners, automated payment by credit
Stiroh, Kevin, J. “Information Technology and the U.S. Productivity Revival” What do the Industry Data
Say?” American Economic Review, December 2002, 92(5), pp.1559-1575.
3 Wall St. Journal, June 17, 2004. P1
and debit cards, and capture of all the sales information on a real-time basis. This and
related company inventory and supply chain management systems have allowed
automated inventory replenishment, sometimes communicated directly to suppliers, and
real-time information on what is selling and what is not. To quote from a recent Wall St.
Journal front page:
…today big retailers know what is selling at each of their stores every day
by the hour. So they don’t have to rely on suppliers to tell them how
much to stock.
As stores have improved inventory controls they have also been better
able to cut costs and lower prices. The lower prices, in turn, helped spur
Americans to keep shopping through the recent recession. 3
These point-of-payment systems have also permitted major productivity gains in other
related areas like automated toll payment and gasoline station transactions.
The impact of so-called E-tailing has also been widespread. U.S.-reported data below
show that E-tailing has been growing rapidly in percentage terms, but is still at only a
modest 5.2% of total retailing.
Figure 18: E Commerce Share of Total Retailing Rising
2000 2001 2002 2003
Ecomm Share of Total Retailing (left scale)
Ecomm Year over Year % change (right scale)
Still, the total is moving into the noticeable area, at $72 billion per year. The recent 25–
30%-per-year growth is showing no sign of slacking, suggesting that share can rise still
further in the next few years. E-tailing now accounts for about 40% of all the recent
growth in retailing.
There are other benefits to Web-based shopping. Consumers can now easily access a
wealth of supporting information free, even if the final purchase is on site or by phone.
So, the data above, tracking Web payment only, are undercounting this. The “free” part is
great for consumers, but not so good for the “dot-com” E-tailers. It turns out that with
heavy competition and a near-zero cost of additional consumer information access, the
permanent long-run competitive price of this valuable service is near zero, or zero with
This was, to state the obvious, a disappointment to investors during the dot-com bust.
However, there is a lesson here. The tremendous value that investors foresaw for the
consumer is actually there—we just do not need to pay for it in most cases. The Web
infrastructure and E-tailing knowledge created during the investment boom is there
permanently for all of us to enjoy.
In limited cases, some value can be captured by firms. The Web auction site eBay, a
recently IPOd venture-supported firm, is a good example. This extra market-making
function can be charged for, and also gives good information to customers—it is fair to
say that such businesses as purchasing antiques have been revolutionized by this.
Customer search costs are a fraction of what they were ten years ago, and information on
what is available is now nationwide, or even worldwide has been multiplied.
Without high tech, retailing sales productivity would be less and prices higher. Our
model scenario suggests that with no high-tech investment boom in the 1990s, retail
prices would be much higher by now. The unfavourable overall productivity implications
are noted later.
The airline industry and airline travel have been transformed in the last ten years. First
comes their essential equipment: the aircraft. There are major advances, like modern
collision avoidance. The aircraft manufacturing industry is also applying unique advances
in Computer-Aided Design and Manufacturing (CAD/CAM).
The advances in airline travel service technology and the investments that support that
have been even more striking. Modern route planning actually requires a lot of computer
power to optimise. The Sabre airline information system permits schedule changes,
customer tracking, bar code luggage tracking, and real-time seat assignment. While air
travel is often no picnic, especially with the new security requirements, the new
technology also permits a level of security that would be unthinkable 20 or even 10 years
ago, including high-tech scanners and information sharing on potential terrorists.
Web-based travel booking has also produced dramatic changes. The Web is a natural for
air travel, and various E-travel services—some venture supported—are proving
economically viable, as well as offering valuable service.
Manufacturing was (and still is) a dominant user of IT investment, and is clearly the
sector with the largest productivity gains. A comparison of actual manufacturing
productivity with results under no high-tech gains scenario are shown below.
Manufacturing Productivity (in levels)
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004
No High Tech Boom Baseline
In the real world, manufacturing productivity showed dramatic gains of 4.2% per year
since 1993, well above the overall productivity gains. Also shown on the chart above, as
our model estimates without hi-tech investment growth, the growth would have been
much less, only 0.8% per year on our calculation. The dominant share of manufacturing
productivity growth (81%) was due to the extra hi-tech investment.
Individual high-tech applications are as diverse as the sector itself. The CAD/CAM noted
earlier allows specialized short runs of different items at low cost. Bar coding, route
planning allows just-in-time inventory and other supply chain management practices that
are similar in some ways to the retailing sector, as noted above. Automated machine tools
and process control are productivity-enhancing investments unique to the goods-
producing industries. It is fair to say that U.S. industries have led the world in the
application of high tech, even though user industries anywhere in the world can buy and
use the equipment.
Total U.S. Economy Benefits
What did high-tech mean for the United States the last ten years? We “shocked” the
Global Insight U.S. Macro model over recent history, since 1993, with the assumption
that dollar computer and telcom equipment sales (the purchase by each user industry)
stayed constant over the last ten years. With a reduced assumption of price decline, “real”
growth in these two categories was held at 3% per year for computers and 1% per year
for telcoms. In reality, they grew 12.1% and 5.2% per year, respectively. These are fairly
limited assumptions about “no high tech.” They still assume small real growth in
purchases. If new products and applications had ground to a halt, there might even have
been large outright declines in the annual outlays.
Total national (private nonfarm) productivity in this scenario continues to grow, but at
reduced rates. The trend of productivity in index levels is shown below.
Nonfarm productivity (in levels)
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004
No High Tech Boom Baseline
In the cumulative period since 1993, productivity is 9.1% higher than the high-tech
boom, or about 0.9% per year. This may not seem like a huge amount, but it is most of
the acceleration noted earlier in the Meyer study for the latest high-tech boom (he
measured from a 1995 date, and the comparison also holds for that period). This can also
make a substantial difference over time.
The following chart shows levels of real GDP per capita in dollars for the history and
Figure 21: Real GDP per Capita (in Levels)
Per capita GDP
1990 1992 1994 1996 1998 2000 2002 2004
No High-Tech Boom Baseline
As indicated, the GDP per capita difference in any area was only about 1% or less by the
end of the ten-year simulation. U.S. GDP per capita, a good measure of the standard of
living, was 7% lower in the simulation, compared with history. Clearly, the high-tech
surge was good for economic progress and the overall standard of living.
Will the Productivity Surge Continue?
This is a good question, and central to how well the country as a whole will perform
economically. However, there are naysayers, including Steven Roach of Morgan Stanley,
who argues that the productivity surge is illusory because it is being badly measured.
However, the dominant academic school of thought, including Stiroh and Jorgenson
noted earlier, Bradford DeLong of Berkeley, and former Treasury Secretary Lawrence
Summers (the current president of Harvard), argue that productivity will surge for at least
several more years. Delong 3writes:
Will this new, higher level of productivity growth persist? The answer
appears likely to be “yes.” The most standard of simple applicable growth
models…predicts that the social return to information technology
investment would have to suddenly and discontinuously drop to zero for
the upward jump in productivity growth to reverse itself in the near future.
More complicated models that focus in more detail on the determinants of
investment spending or on the sources of increased total factor
productivity appear to strengthen, not weaken, forecasts of productivity
growth over the next decade.
To give some examples, the spreadsheet, the World Wide Web, and Amazon.com were
all innovations made possible and accessible to the masses by cheaper computing prices.
Delong argues that the pace of innovation will continue into the foreseeable future
because of the still-declining price of information technology. The latest data showing the
sustained R&D effort and continuing VC-related activity, even if the latter is at post-
boom levels, also points to a continuing flow of productivity-enhancing innovations.
In 2002, productivity growth surged 5.3%, the highest performance in the past 50 years.
It slowed only modestly in 2003 and 2004, estimated at 4.4% and 4.0%, respectively.
These amazing numbers indicate that the surge in productivity that started in the 1990s is
not only continuing but also accelerating. Some of this recent productivity came from an
austerity-related temporary reluctance to hire workers, and a workoff of earlier excess
hiring. However, even this was partly enabled by the ongoing application of technology.
However, some slowing appears in the latest data. Some of the exceptional austerity-
driven productivity of the last three years is moderating.
In Global Insight’s latest long-term forecast (May 2004), productivity growth averages
2.6% in 2003–14, still well above historical averages. This forecast assumes that some of
the temporary austerity-related sources will fade, but the permanent long-term benefits to
productivity coming from information technology and medical gains, and the venture
activity that helped to give birth to them, will persist for another ten years.
3 Productivity Growth in the
J. Bradford DeLong1
University of California at Berkeley and NBER
A Historic View of Innovation
This section examines key elements of how U.S. innovation has proceeded in history. We
examine how new technological sectors evolved in the pre-WW II period through a
process of “creative destruction.” We also look at the pattern of formation of the biggest
U.S. Fortune 500 industrial firms over the U.S. business cycle since 1906.
The business sectors examined—automobiles, radio, and aircraft— were all technological
leaders in their day, and all had a massive early expansion of new small companies, each
attempting to find the best business model to apply to the new technology. What followed
was a massive shakeout, as most of the new, small firms either went bankrupt or were
merged and reorganized.
The term creative destruction was coined by the famous economist Joseph Schumpeter in
his work “Capitalism, Socialism and Democracy.” He described a process. “the gale of
creative destruction.” where new dynamic firms eventually drove their competitors out of
business. Phrases he used included “industrial mutation…that incessantly revolutionizes
the economic structure from within.” He cited “steel…from the charcoal furnace to our
own type of furnace” and “transportation from the mail coach to the airplane.” The
following are detailed looks at several sectors in history that best illustrate this. Of
course, the process is still under way, as seen in the recent shakeouts of PC-producing
and Internet companies.
The great early pioneers of automobile technology often had companies and models
named after them. According to the Antique Automobile Club of America1
(http://www.aaca.org/autohistory/01.html), it was the “original mechanical wizards like
Henry Ford, Ransom E. Olds, David Dunbar Buick, William Knudsen, Henry Leland,
Charles Kettering and the Dodge brothers (as well as the Packard brothers and foreign
inventors like Mercedes, Benz and Peuguot) who invented and figured out how to build
the automobile.” Most of these founders formed their first companies before 1900.
Between 1900 and 1910, there were 290 identified automobile companies formed, and no
doubt many more uncounted. Of note, Henry Ford, fired from the bankrupt Detroit Auto
Company, formed the present Ford Motor Company in 1901. Bankruptcies,
reorganizations, and mergers were common during this period. The automobile assembly
line was invented by Ford in 1913. The subsequent growth of the Model T's market share
to over 75% marked the beginning of major consolidation in the industry forced by the
economies of scale. Technology advanced rapidly, with balloon tires, batteries, and
starters being key features that triggered smaller innovation surges in auto supply
By 1926, there were 43 auto companies. No more U.S. entrants were formed after that.
By 1929, the big three—GM, Ford, and Chrysler—had consolidated at least 80% of U.S.
auto sales. By 1939 and the end of the great recession, the big three had at least 90% of
the market. Of the remaining 10%, most was Hudson, Nash, Packard, Studebaker, and
Willys—names we remember that are no longer around—which later merged under
pressure. So the recent upheaval, consolidations, and major bankruptcies of the dot-coms,
telcom, fiber optic, and other new technologies are not really new. Human ingenuity is
not at a level where it can choose the best path and technology ahead of time.
Radio is notable for having gone through at least two separate creative destruction cycles.
In the first stage, wireless telegraphy, invented by Marconi in 1895, went through a huge
cycle of excess investment in sending and receiving stations, especially for maritime and
cross-ocean messages. From Modern Electronics, December 1911:
“The commercial stations communicate without difficulty to ranges in excess of five
hundred miles. Commercial stations as well as government ones, dot the coasts from one
boundary to the other, sending countless messages and safeguarding the lives of
thousands of human beings on ship-board. It is with deep regret that a number of
unscrupulous men should have gained a footing in the wireless industry, and used it as a
tool for extorting money from thousands of victims in promoting stock sales. This money
has been used, not for wireless stations and service, but for the foundation of huge
fortunes made by these men in their dishonest undertaking. Today it may be said that the
wireless industry is undergoing a state of purification, and will issue from this state, a
clean and prosperous organization, its main financial income being secured as tolls for
messages. The promoting stage is one to be passed by every new invention, and if that
industry survives through this period of criticisms and abuse, its future success is
Some things don’t change. At least most of the modern dot-coms, fiber optic, and other
recent failed ventures thought they would have a viable business and failed for other
The second radio boom was for broadcast sound, including the equipment, and “content.”
From Oscar Lescarboura, Radio for Everyone, 1922:
There are various other organizations devoting a goodly part of their
efforts to broadcasting radio-phone news and concerts. In fact, as things
stand at present it is safe to state here that virtually every part of the
United States is covered by one or more stations. To give a list of stations
is virtually impossible, for in an art that is so new there are bound to be
frequent changes. Hence no attempt is being made to offer a list, because
it would be hopelessly obsolete by the time it got into print. The reader is
referred to the radio periodicals and to the daily newspapers that have
radio sections, for the last-minute information on radio-phone stations.
Consolidation began almost immediately. Somewhat as today with dot-coms, there was a
struggle to find a profitable business model. There was even a contest with a prize for the
best money-making business proposal. Like today, as described earlier for the dot-coms,
the pricing model turned out to be quite similar—“free with advertising.”
RCA (spun off from General Electric) and Westinghouse moved quickly to consolidate
both radio broadcast networks and production of receivers, and quickly became dominant
in both these areas. Thousands of broadcast stations went out of business, particularly
after licensing and laws controlling the airwaves were passed.
Following the first controlled, powered flight by the Wright Bros, the industry expanded
rapidly. The Curtiss company achieved early dominance, and eventually merged with the
Wright’s. Other planebuilders we recognize also went into business before 1920,
including Douglas, Boeing, Loughead (Lockheed), and Martin. There were at least ten
other companies that never achieved the recognition stage, like Gallodet and Thomas
Bros, that were merged or went out of business.
Postwar WWI brought many new names, including Convair, Northrup, and Hughes, and
a lot more, at least 20, that merged or went out of business, like Consolidated Vulta.
There was first airmail and then the emergence of scheduled airlines in the 1930s that
fueled new civilian aircraft interest. Civilian aircraft production by then had seen some
consolidation down to Boeing, Douglas, and Ford (the auto company) as major suppliers.
WWII brought a massive demand for warplanes. More than 160,000 were produced by
war’s end. By then, the largest companies were:
• Boeing: B-17, B-29 bombers.
• Convair: B-24 bomber.
• Lockheed: P-38 fighter.
• Curtiss: P-40 fighter, C-46 transport.
• Douglas: C-47, C-54 transports.
• North American: P-51 fighter.
• Republic: P-47 fighter.
While jet aviation created new technological marvels and commercial interest, U.S.
civilian aircraft companies were by then dominated by Boeing and McDonnell Douglas.
Foreign competition had always been strong, and the U.S. civilian aircraft market saw
more foreign supply, including Airbus, consolidated in the 1970. Boeing and Airbus, with
some smaller suppliers like the Brazilian Embraer, are the only survivors of the global
civilian aircraft market. The military market continued to expand, and the technological
lead fell to aerospace and the massive U.S. NASA and military space and missile effort.
Companies Created in the Down-cycle Years of the U.S. Economy
This section includes a number of Fortune 500 companies listed in the down-cycle years.
It also compares the number of companies listed in the up-cycle and down-cycle years
and gives an overall picture of the economy.
Figure 22 depicts the trend since 1906 of companies in the United States filing their
initial public offerings (IPOs). The years highlighted in orange show the companies
created in down-cycle years of recession in the U.S. economy.
Figure 22: Identified IPOs of Fortune 500s Since 1900
Number of Companies IPO'd
Of the 237 Fortune 500 companies with identified IPOs since 1906, 207 companies filed
their IPOs in up-cycle years, and the remaining 30 in downcycle years
Prior to World War II, there was a small number of Fortune 500 companies with
identified IPO dates. While thousands of companies were created during this period, only
21 are in the current Fortune 500 list. Of these, seven were created in pre-war down-cycle
years. The 1930s were particularly depressed, and only four companies were created.
They were Masco, Northrop Grumman, PepsiAmericas, and Parker Hannifin.
Postwar, there was a comparatively small business cycle link until 1980. The 1970s were
a particularly dry period. After 1980, business picked up, with a lull in 1988–91. It is
interesting to note that more recently, 66 Fortune 500 companies had IPOs in the 1990s.
Well over half of these were venture-supported companies, but most filed their IPOs in
the up-cycle years. The IPO trend followed a pattern similar to that followed by the U.S.
As elaborated on in the next chapter for recent IPO history, the formation of these
companies was quite cyclical, driven by the U.S. business and financial cycles.
Nevertheless, once born, they proceeded to expand and mature. They ended up as the
biggest U.S. companies, with relatively stable performances that are only affected in a
minor way by the fluctuations in the financial markets and economy.
Venture Investing: a Volatile Input to Business
This chapter discusses the volatility of venture activities and returns and stock markets, a
troublesome feature of all free markets. We conclude that while the early stage VC IPOs
and the financial markets are all highly volatile, the later, more mature stages of
expansion become much less volatile. In the final mature stages, we note that new
spending investment and stock market growth on capital equipment and the key high-tech
sector still shows some cycle in new spending. However, the U.S. capital stock, both
total business plant and equipment and the high-tech part, has continued to display an
upward growth pattern. This embodiment of all the usable plant and equipment we have
has never had a down year for growth, though the positive growth rate shows some
VC Investments with Respect to the Business Cycle
Starting at the very beginning, there are six early stages in the investment financing of a
firm: seed, startup, expansion, mezzanine, buyout, and (if needed) turnaround. Most
venture outlays focus on the seed, startup, and expansion stages.
Let us look at so-called seed activities. A tiny fraction of VC money, about 2%, goes in
earliest-stage financing, called seed money, which constitutes funds for initial research to
prove a concept. A significant proportion of VC is invested to support product
development and initial marketing (often referred to as startup funds).
The figure below shows the VC disbursement in startup/seed activities:
Figure 23: VC Investment in Startup/Seed Activities
VC investment (in $ million)
Source: Thomson Venture Economics/National Venture Capital Association
In 1980–2002, seed/startup activities constituted $21.4 billion out of the total US$339.9
billion invested in all the business stages, accounting for approximately 6.3% of all U.S.
VC disbursements. Seed/startup activities rose from $157.5 million in 1980 to a first peak
of $1.5 billion in 1986, a nearly ten-fold increase. They then fell to $241 million in 1991,
for an 83.9% decline. Seed/early money then ramped up to a peak of $3.3 billion in 1999,
leading the high-tech (and medical) boom and other sectors as well. The latest contraction
was also dramatic: a 90% decline from 1999 to a low of $352 million in 2002. It
remained roughly the same last year, at $354 million.
This early cycle may be driven by funds availability and optimism or pessimism.
However, it also may reflect how many promising ideas have been generated at that point
by ongoing innovation and the advance of knowledge. The early seed cycle will also
partly drive the later cycles.
The cycle for overall venture placements is also highly volatile, as shown in the table
below. The following figure shows the percentage change in the value of VC and VC-
backed IPOs, compared with the previous year, in 1983–2003.
Figure 24: High Volatility for Total VC and VC-Backed IPOs
Percent change from a year earlier
1981 1985 1989 1993 1997 2001
Venture Investment ($ millions) Number of IPOs
IPO Offer Amount ($ millions)
Source: Thomson Venture Economics/National Venture Capital Association
Total VC placements jumped at growth rates exceeding 100% in 1981 and 1999, but
demonstrated less volatility than the seed/startup cycle. The two VC cycles look
somewhat similar, especially the declines in 1989–91, partly because seed/startup drives
the later infusions. IPO activity has been even more volatile in both numbers and total
value. Note that in percentage terms, the early cycles were about as severe as the 2000–
02 boom bust in both upturns and downturns.
The following chart shows the number of VC-supported and non-VC IPOs since 1992.