[FRIAM] HBR: Breakthrough Ideas for 2007 (fwd)

[Richard Lowenberg]

From the article by G. West.     rl


11. Innovation and Growth: Size Matters *

Executives talk about their companies’ “DNA” and roles in “business
ecosystems,” but the analogy to living organisms is more than
metaphorical. Like the mathematical laws governing how organisms’
metabolism, growth, evolution, and mortality depend on size, there
are rules that appear to govern the growth, performance, and even
decline of cities and other social organizations. Although we can’t
yet predict how specific cities or companies will evolve, we’ve found
general mathematical relationships between population size,
innovation, and wealth creation that may have important implications
for growth strategy in organizations.

In biology, different species are in many ways scaled versions of one
another. Bacteria, mice, elephants, sequoias, and blue whales may
look different, but most of their fundamental characteristics,
including energy and resource use, genome length, and life span,
follow simple mathematical rules. These take the form of so-called
power-law scaling relationships that determine how such
characteristics change with size. For example, metabolic rate
increases as the ¾ power of mass. Put simply, the scaling law says
that if an organism’s mass increases by a factor of 10,000 (four
orders of magnitude), its metabolic rate will increase by a factor of
only 1,000 (three orders of magnitude). This represents an enormous
economy of scale: the bigger the creature, the less energy per pound
it requires to stay alive. This increase of efficiency with size—
manifested by the scaling exponent ¾, which we say is “sublinear”
because it’s less than one—permeates biology. These ubiquitous
scaling laws have their origin in the universal properties of the
networks that sustain life, such as the cardiovascular and
respiratory systems.

Social organizations, like biological organisms, consume energy and
resources, depend on networks for the flow of information and
materials, and produce artifacts and waste. So it would not be
surprising if they obeyed scaling laws governing their growth and
evolution. Such laws would suggest that New York, Santa Fe, New
Delhi, and ancient Rome are scaled versions of one another in
fundamental ways—as, potentially, are Microsoft, Caterpillar, Tesco,
and Pan Am. To discover these scaling laws, Luís Bettencourt at Los
Alamos National Laboratory, José Lobo at Arizona State University,
Dirk Helbing at TU Dresden, and I gathered data across many urban
systems in different countries and at different times, addressing a
wide range of characteristics including energy consumption, economic
activity, demographics, infrastructure, intellectual innovation,
employment of “supercreative” people, and patterns of human behavior
such as crime rates and rates of disease spread.

We did indeed find that cities manifest power-law scaling similar to
the economy-of-scale relationships observed in biology: a doubling of
population requires less than a doubling of certain resources. The
material infrastructure that is analogous to biological transport
networks—gas stations, lengths of electrical cable, miles of road
surface—consistently exhibits sublinear scaling with population.

However, to our surprise, a new scaling phenomenon appeared when we
examined quantities that are essentially social in nature and have no
simple analogue in biology—those associated with innovation and
wealth creation. They include patent activity, number of
supercreative people, wages, and GDP. For such quantities the
exponent (the analogue of ¾ in metabolic rate) exceeds 1, clustering
around a common value of 1.2. Thus, a doubling of population is
accompanied by more than a doubling of creative and economic output.
We call this phenomenon “superlinear” scaling: by almost any measure,
the larger a city’s population, the greater the innovation and wealth
creation per person.

By almost any measure, the larger a city’s population, the greater
the innovation and wealth creation per person.

Organismic growth, constrained by sublinear power-law scaling derived
from the dynamics of biological networks, ultimately ceases, with the
equations predicting what size organisms will reach. In contrast, our
equations predict that growth associated with superlinear scaling
processes observed in social organizations is theoretically
unbounded. This would seem to bode well for organizations.
Unfortunately, however, the equations also predict that in the
absence of continual major innovations, organizations will stop
growing and may even contract, leading to either stagnation or
ultimate collapse. Furthermore, to prevent this, the time between
innovations (the “innovation cycle”) must decrease as the system grows.

Though our research has focused on cities, the social and structural
similarities between cities and firms suggest that our conclusions
extend to companies and industries. If so, the existence of
superlinear scaling that links size and creative output has two
important consequences: First, it challenges the conventional wisdom
that smaller innovation functions are more inventive, and perhaps
explains why few organizations have ever matched the creativity of a
giant like Bell Labs in its heyday. Second, it shows that because
organizations and industries must apparently innovate at a
continually accelerating rate to avoid stagnation, economizing by
reflexively cutting R&D budgets and creative staffs may be a
dangerous strategy over the long term.

Geoffrey B. West (gbw at santafe.edu) is the president of the Santa Fe
Institute in Santa Fe, New Mexico.