Better ideas take over in the 1930s, giving everyone a free lunch

General Motors Futurama visitors at the 1939-1940 New York World’s FairGeneral Motors Futurama visitors
at the 1939-1940 New York World’s Fair [1]

When better ideas take over, they add more value than the sum of the capital and labor inputs

A rough measure of the technological and organizational progressivity of an era is how much more rapidly output grows than a weighted average of the growth of labor and physical capital (structures and equipment). That difference, or residual, represents the growth of total factor productivity.

This chapter examines the years 1929-41 in the United States…

…output rose very rapidly… almost entirely as the consequence of the growth of total factor productivity.

Total Factor Productivity  annual growth rates in the US for 1873-2010

Double digit unemployment for more than a decade represented a terrible waste of human and other resources, and untold hardship for the millions of people out of work.

And yet the Depression years were also a triumph of American ingenuity, inventiveness, and hard work. What I have called the country’s Great Leap Forward…

Better ideas take over in the 1930s in processes and materials

…the 1930s were… a great age of process and materials innovations.

  • The 1930s… saw the tail end of the revolution in factory layout and design that had produced such extraordinary TFP gains in manufacturing between 1919 and 1929 (over five percent per year). That revolution involved replacing systems for distributing power internally within a factory.
  • Machinery became larger… …capital and operating costs per unit of output dropped when capacity increased.
  • …topping techniques used the steam from high pressure boilers to heat lower pressure boilers.
  • More generally, throughout industry, exhaust gasses from stacks were used to preheat air to improve combustion, preheat materials for subsequent fabrication, or generate steam…
  • Improvements in thermal efficiency also benefited from attention to low cost but often high payoff investments in insulation.
  • Similarly, modest investments in instrumentation yielded big efficiency gains, facilitating automatic process control, which lengthened the life of equipment, and reduced downtime and maintenance costs.
  • Better treatments extended the life of wooden railroad ties from eight to twenty years.
  • Stainless steel reduced oxidization on railway cars, while chrome plating lengthened the lives of tools and moving parts. Carbon steel blades had to be removed and resharpened after cutting 60 feet of plastic. A tungsten carbon alloy blade could cut 10,000 feet without refitting.
  • Quick drying lacquers reduced the time needed to paint a car from more than three weeks in the early 1920s to a few hours, with consequent reductions in inventory costs.
  • During the 1930s mechanical refrigerators moved from a “bleeding edge” product to a mass production and mass consumption item.
  • In 1928 the Dupont Corporation lured Wallace Caruthers away from his laboratory at Harvard to Delaware, where he began to develop blockbuster new materials including neoprene and nylon. Caruthers… committed suicide in 1937…

Better ideas take over in the 1930s in transportation

The other major source… was spillovers in transportation and distribution resulting from the  buildout of the surface road network.

  • Railroad productivity soared, in part because of institutional and organizational changes involving freight interchange…
  • …there was enough production during the Depression to replace every truck and bus on the road in 1929 at least once, and add millions more to the transportation system. These newer vehicles were, on average, larger, more powerful, and more reliable.
  • …cars were much improved. Radios, heaters, and four wheel hydraulic brakes were now standard. Automatic transmission, power steering, and more powerful engines became options. Tires moved from the narrow profile high pressure products of the 1920s – reflecting the birth of the automobile in the bicycle industry, to the low pressure balloon tires upon which most of us roll today. Vehicles were streamlined and more aerodynamic, with headlights and trunks incorporated into the body rather than addons.
  • In 1936 Donald Douglas introduced the DC3 – arguably the world’s most famous and successful aircraft (it had a starring role in the closing scenes of the movie Casablanca, alongside Humphrey Bogart and Claude Rains). …all US aircraft that saw major service operation in World War II were already on the drawing boards (“substantially designed”) in December of 1941…[2]

  1. Field, Alexander J. A great leap forward: 1930s depression and US economic growth. Yale University Press, 2011.
  2. Field, Alexander J. “Economic Growth and Recovery in the United States: 1919–1941.” The Great Depression of the 1930s: Lessons for Today, edited by Nicholas Crafts and Peter Fearon, Oxford University Press, 2013, pp. 358-394.

Learn well by high use of practice testing and distributed practice

Comparison of learning techniques showing that practice testing and distributed practice are the keys to learn well

Learn well by high use of high-utility techniques

Practice testing

For example… practicing recall of target information via the use of actual or virtual flashcards, completing practice problems or questions included at the end of textbook chapters, or completing practice tests included in the electronic supplemental materials that increasingly accompany textbooks.

Distributed practice

…distributing learning over time (either within a single study session or across sessions)… we use the term distributed practice to encompass both spacing effects (i.e., the advantage of spaced over massed practice) and lag effects (i.e., the advantage of spacing with longer lags over spacing with shorter lags)…

Learn well by moderate use of moderate-utility techniques


…having students explain some aspect of their processing during learning.

Interleaved practice

…students alternate their practice of different kinds of items or problems.

Elaborative interrogation

…prompting learners to generate an explanation for an explicitly stated fact.

Learn well by low use of low-utility techniques

The keyword mnemonic

Imagine a student struggling to learn French vocabulary, including words such as la dent (tooth), la clef (key), revenir (to come back), and mourir (to die). …the student would first find an English word that sounds similar to the foreign cue word, such as dentist for “la dent” or cliff for “la clef.” The student would then develop a mental image of the English keyword interacting with the English translation. So, for la dent–tooth, the student might imagine a dentist holding a large molar with a pair of pliers.

Imagery use for text learning

Students…were told to read the text and to mentally imagine the content of each paragraph using simple and clear mental images.


…having students write summaries of to-be-learned texts. Successful summaries identify the main points of a text and capture the gist of it while excluding unimportant or repetitive material…


…rereading differentially affects the processing of higher-level and lower-level information within a text, with particular emphasis placed on the conceptual organization and processing of main ideas during rereading.

Highlighting and underlining

…students report… underlining, highlighting, or otherwise marking material as they try to learn it…[1]

  1. Dunlosky, John, et al. “Improving students’ learning with effective learning techniques: Promising directions from cognitive and educational psychology.” Psychological Science in the Public Interest 14.1 (2013): 4-58.

Self-care in marriage makes partners more positive, happy, well, and connected

Ackerman's PPIK theory of intelligence as Process, Personality, Interests, and intelligence as Knowledge illustrates the course of self-care, which, in marriage, makes partners more happy, well, positive, and connected

gp = intelligence-as-process
gk = intelligence-as-knowledge
R = Realistic interests
A = Artistic interests
I = Investigative interests
TIE = Typical Intellectual Engagement
gf = fluid intelligence
gc = crystallized intelligence

Illustration of Ackerman’s PPIK theory, outlining the influences of intelligence-as-Process, Personality, Interests, and intelligence-as-Knowledge during adult development, covering academic and occupational knowledge.

The representation indicates that measured fluid intelligence (Gf) develops out of intelligence-as-process (gp), and that measured crystallized intelligence (Gc) develops out of (or is a consequence of) intelligence-as-knowledge (gk).

Interests (Realistic, Investigative, and Artistic) and personality traits (Openness and TIE) are influenced by intelligence to some degree, and in turn, influence knowledge.

Self-care in marriage keeps partners taking small steps that add up

…many intellectually demanding tasks in the real world cannot be accomplished without a vast repertoire of declarative knowledge and procedural skills. The brightest (in terms of IQ) novice would not be expected to fare well when performing cardiovascular surgery in comparison to the middle-aged expert, just as the best entering college student cannot be expected to deliver a flawless doctoral thesis defense, in comparison to the same student after several years of academic study and empirical research experience. In this view, knowledge does not compensate for a declining adult intelligence; it is intelligence!

Moreover, the importance of personality and interests as determinants of the direction and amount of effort expended in the acquisition and maintenance of intelligence-as-knowledge should not be underestimated. Small correlations at the micro-level, when aggregated as influence over time…, may help us predict and understand why some adults continue to acquire knowledge in particular areas and others do not.[1]

Self-care in marriage lets partners build optimism 

Explanatory style is the habitual way in which people explain the bad events that befall them… Three dimensions of these explanations are of interest: stability versus instability, globality versus specificity, and internality versus externality.

A stable cause invokes a long-lasting factor (“it’s never going to go away”), whereas an unstable cause is transient (“it was a one-time thing”).

A global cause is one that affects a wide domain of activities (“it’s going to ruin everything I do”), whereas a specific cause is circumscribed (“it has no bearing on my life”).

Finally, an internal cause points to something about the self (“it’s me”), whereas an external cause points to other people or circumstances (“it’s the heat in this place”).

Pessimistic explanatory style (the belief that bad events are caused by stable, global, and internal factors) predicted poor health at ages 45 through 60, even when physical and mental health at age 25 were controlled. Pessimism in early adulthood appears to be a risk factor for poor health in middle and late adulthood.[2]

Self-care in marriage makes partners more happy and well

The seven protective factors that distinguish the happy-well from the sad-sick are under at least some personal control.

Self-care increases happiness and wellness

Odds ratios of happy-well to sad-sick or dead 
Variable College men age 75-80 Core-city men age 65-70
No alcohol abuse very high 4.56 to 1
Without depressive diagnosis 10.4 to 1 3.51 to 1
Smoking <30 pack-years 4.81 to 1 4.56 to 1
Some regular exercise 3.09 to 1 unknown
Body mass index >21 and <29 3.05 to 1 1.71 to 1
Mature defenses 2.65 to 1 2.98 to 1
Stable marriage 1.94 to 1 2.75 to 1
Parental social class 1.46 to 1 1.12 to 1
Education unknown 0.86 to 1
Ancestral longevity 1.00 to 1 1.00 to 1
Warmth of childhood 0.98 to 1 0.99 to 1
Childhood temperament 0.92 to 1 1.10 to 1


Self-care in marriage makes partners more positive and connected

To be well psychologically is more than to be free of distress or other mental problems. It is to possess positive self-regard, mastery, autonomy, positive relationships with other people, a sense of purposefulness and meaning in life, and feelings of continued growth and development.[4]

  1. Ackerman, Phillip L. “Domain-Specific Knowledge as the “Dark Matter” of Adult Intelligence: Gf/Gc, Personality and Interest Correlates.Journal of Gerontology: Psychological Sciences 55.2 (2000): P69-P84.
  2. Peterson, Christopher, Martin E. P. Seligman, and George E. Vaillant. “Pessimistic Explanatory Style Is a Risk Factor for Physical Illness: A Thirty-Five-Year Longitudinal Study.Journal of Personality and Social Psychology 55.1 (1988): 23-27.
  3. Vaillant, George E., and Kenneth Mukamal. “Successful Aging.American Journal of Psychiatry 158.6 (2001): 839-847.
  4. Ryff, Carol D. “Psychological Well-Being in Adult Life.Current Directions in Psychological Science 4.4 (1995): 99-104.

Learn fast by pursuing skills, challenges, and rapid feedback

Four girls learn fast on computer together.

Learn fast by building skills

…determining what a student should be able to do is far more effective than determining what that student should know. It then turns out that the knowing part comes along for the ride.

The objectives should be skills, not knowledge. …understand as much as possible about the mental representations that experts use, and teach the skill so as to help students develop similar mental representations.

…you don’t build mental representations by thinking about something; you build them by trying to do something, failing, revising, and trying again, over and over.

Learn fast by taking on challenges

This will involve teaching the skill step by step, with each step designed to keep students out of their comfort zone but not so far out that they cannot master that step.

To help… physics students… develop… mental representations, Wieman and his coworkers developed sets of clicker questions and learning tasks…

The clicker questions and tasks were chosen to trigger discussions that would lead the students to grapple with and apply the concepts they were learning and, ultimately, to those concepts to answer the questions and solve the tasks.

The questions and tasks were also designed to push the students outside their comfort zones—to ask them questions whose answers they’d have to struggle for—but not so far outside their comfort zones that they wouldn’t know how to start answering them.

Wieman and his colleagues pretested the clicker questions and learning tasks on a couple of student volunteers who were enrolled in the course. They gave these students the questions and the learning and then had them think aloud as they reasoned their way toward the answers.

Based on what the researchers heard during the think-aloud sessions. they modified the questions and tasks, with a specific emphasis on avoiding misunderstandings and questions that were too difficult for the students to deal with. Then they went through a second round of testing with another volunteer, sharpening the questions and learning tasks even more.

Learn fast by pursuing rapid feedback

…the classes were structured so that the students would have the opportunity to deal with the various concepts over and over again, getting feedback that identified their mistakes and showed how to correct them. Some of the feedback came from fellow students in the discussion groups and some from the instructors, but the important thing was that the students were getting immediate responses that told them when they were doing something wrong and how to fix it.

…the regular cycle of try, fail, get feedback, try again, and so on is how the students… build their mental representations.[2]

  1. “Clickers in the science classroom (and you don’t even need the clickers).”, 22 Jan. 2015, Accessed 21 Dec. 2016.
  2. Ericsson, Anders, and Robert Pool. Peak: Secrets from the New Science of Expertise. Houghton Mifflin Harcourt, 2016, pp. 250-253. See also:
    Deslauriers, Louis, Ellen Schelew, and Carl Wieman. “Improved learning in a large-enrollment physics class.” Science 332.6031 (2011): 862-864.

English-language networks bring a better quality of life

Per-capita income vs. English Proficiency Index score, showing rising impact of English-language networks







Low Moderate High


Where English skills are very high (where EF EPI Scores are 63.2 or higher), per-capita income jumps above the trend line, or well above.

English-language networks open up research

English serves as a bridge that connects employees across countries and cultures, weaving networks for innovation.

By a wide margin, researchers in the United States publish the most scientific papers every year, and the United Kingdom ranks third in publication numbers, after China. However,… Chinese research accounts for only 4% of global citations in science publications, compared to 30% for the U.S. and 8% for the U.K. …Chinese research is less integrated into the global knowledge economy.

English skills allow innovators to read primary scientific research, form international collaborations, bring in talent from overseas, and participate in conferences. English proficiency expands the number of connections innovators can make with the ideas and people they need to generate original work.

English-language networks open up business

English spread as a language of international trade and diplomacy first under the British Empire, and then during the postwar economic expansion of the United States.

An increasing number of companies headquartered in non-English-speaking countries (e.g., Rakuten, Renault, and Samsung) have adopted English as their corporate language.

Networks bring better salaries and quality of life

An improvement in English proficiency is tied to a rise in salaries… In many countries, higher English proficiency correlates with a lower unemployment rate among young people.

The Human Development Index measures education attainment, life expectancy, literacy, and standards of living. …all High and Very High Proficiency countries are rated “Very High Human Development” on the HDI.

…English is a core skill today. …it should be taught and tested at a level equivalent to native language reading and math skills.

EF EPI methodology

The data for this sixth edition was calculated using results from 950,000 test takers who completed three different EF English tests in 2015. Two tests are open to any Internet user for free. The third is an online placement test used by EF during the enrollment process for English courses.[1]

  1. EF EPI: EF English Proficiency Index. 6th ed., Education First, 2016.

Math notation serves as pointers to both data and procedures

Math notation provides pointers to both data and algorithms all at once.[1]

Math notation works with our limited attention by greatly compressing information

…human brains, though exceedingly complex, are only able to concentrate consciously on a few things at once, requiring a mechanism to cope with the complication:

  • “…early processing is largely parallel – a lot of different activities proceed simultaneously. Then there appear to be one or more stages where there is a bottleneck in information processing. Only one (or a few) ‘object(s)’ can be dealt with at a time. This is done by temporarily filtering out the information coming from the unattended objects. The attentional system then moves fairly rapidly to the next object, and so on, so that attention is largely serial (i.e., attending to one object after another) not highly parallel (as it would be if the system attended to many things at once).”
  • “Mathematics is amazingly compressible: you may struggle a long time, step by step, to work through some process or idea from several approaches. But once you really understand it and have the mental perspective to see it as a whole, there is often a tremendous mental compression. You can file it away, recall it quickly and completely when you need it, and use it as just one step in some other mental process. The insight that goes with this compression is one of the real joys of mathematics.”[2]

Math notation works by being a pointer to both data and procedures at the same time

…process and concept are combined in a single notion… in the working practices of professional mathematicians and all those who are successful in mathematics. They employ the simple device of using the same notation to represent both a process and the product of that process:

  • The symbol 5+4 represents both the process of adding through counting all or counting on and the concept of sum (5+4 is 9).
  • The symbol 4×3 stands for the process of repeated addition “four multiplied by three” which must be carried out to produce the product of four and three which is the number 12.
  • The algebraic symbol 3x+2 stands both for the process “add three times x and two” and for the product of that process, the expression “3x+2”.[1]

  1. Gray, Eddie M., and David O. Tall. “Duality, ambiguity, and flexibility: A ‘proceptual’ view of simple arithmetic.” Journal for Research in Mathematics Education 25.2 (1994): 116-140.
  2. Gray, Eddie, and David Tall. “Abstraction as a natural process of mental compression.” Mathematics Education Research Journal 19.2 (2007): 23-40.

Data collection, sharing, and use are keys at Google

Google meeting with laptops everywhere and video shared online, illustrating Google's data collection, sharing, and use.

Collecting data

“We need generalists… Lots of projects and companies grow without doing new things; they just get bigger teams. We want projects to end.”

Google… tackles most big projects in small, tightly focused teams, setting them up in an instant and breaking them down weeks later without remorse. “Their view is that there is much greater progress if you have many small teams going out at once.”

A typical task, from tweaking page designs to doing scientific research, involves six people. Hundreds of projects go on at the same time. Most teams throw out new software in six weeks or less and look at how users respond hours later.

With 82 million visitors and 2.3 billion searches in a month, Google can try a new user interface or some other wrinkle on just 0.1% of its users and get massive feedback, letting it decide a project’s fate in weeks. One success in ten tries is okay; one in five is superb.

Everyone from a failed venture moves to another urgent project. “If something is successful, you work it in, somehow… If it fails, you leave.”

Sharing data

Google… shares all the information it can with as many employees as possible…

It also pursues a rapid-fire food-fight strategy that throws out ideas as fast as possible, to see what sticks.

One key rule: You can’t call any idea “stupid.”

(Nor is most any idea too wild. On a recent day at the Google campus a bulletin board invited workers to a session on the dream of erecting a 200-mile-high elevator into space.)

Using data

One true god rules at Google: data. The more you collect, the more you know and the more certain your decisions can be, disciples believe…

“Often differences of opinion between smart people are differences of data…”

In some meetings people aren’t allowed to say “I think…” but instead must say “The data suggest…”

…the guy with the best data wins.[2]

  1. “Search Quality Meeting: Spelling for Long Queries (Annotated)” YouTube, 12 Mar. 2012, Accessed 24 Nov. 2016.
  2. Hardy, Quentin. “Google Thinks Small.” Forbes 176.10 (14 Nov. 2005): 198-202.

Print exposure potently builds up cognitive ability


Television… is not the great information machine… …exposure to television… did not predict additional variance over and above… ability…

…when speculating about variables in people’s ecologies that could account for cognitive variability—in an attempt to supplement purely genetic accounts of mental ability…—researchers should find print exposure worth investigating, because such variables must have the requisite potency to perform their theoretical roles. A class of variables that might have such potency would be one that has long-term effects because of its repetitive or cumulative action. Schooling is obviously one such variable… …print exposure is another variable that accumulates over time into enormous individual differences. …these individual differences are associated to a strong degree with individual differences in general knowledge.

…individual differences in declarative knowledge bases—differences emphasized by many contemporary theories of developmental growth—appear to some extent to be experientially based, and the experience that has a particularly close link with these individual differences seems to be print exposure…

Print exposure… determines individual differences in knowledge bases, which in turn influence performance on a variety of basic information processing tasks…

…the more avid readers in our study—independent of their general abilities—knew more about how a carburetor worked, were more likely to know that Vitamin C was in citrus fruits, knew more about how lending rates affect car payments, were more likely to know who their U.S. Senators were, knew more about broiling food, were more likely to know what a stroke was, were more likely to know what a capital-intensive industry was, and were more likely to know who the United States was fighting with and who it was fighting against in World War II.

Print exposure accounted for a sizable portion of variance in measures of general knowledge, even after variance associated with general cognitive ability was partialed out. There does appear to be differential exposure to information, primarily through the medium of reading, and this differential exposure is predictive over and above general cognitive ability. …print exposure… was a more potent predictor than the ability measures. When entered after the print exposure composite, the… cognitive ability measures… accounted for an additional 5.1% of the variance in the knowledge composite scores, considerably less than the 37.1% of the variance accounted for by the print exposure…

…print exposure influences cognitive efficiency.[2]

  1. “Wonder Words… Wisdom, Skill & Virtue.” LifeSuccessCoaching, 30 June 2011, Accessed 20 Nov. 2016.
  2. Stanovich, Keith E., and Anne E. Cunningham. “Where Does Knowledge Come From? Specific Associations between Print Exposure and Information Acquisition.” Journal of Educational Psychology 85.2 (1993): 211-29.

Proven components are ideal for innovation

Diagram showing that fracking used proven components.[1]

knowledge broker… firms innovate by combining existing technologies in new ways that result in dramatic synergy.

Explore new territories

Gain access to a wide range of industries or knowledge domains by working on a wide range of different problems and their existing solutions. The more diverse the experiences of everyone in your organization, the more diverse the set of past solutions to draw upon when facing a new problem.

“Working with companies in such dissimilar industries as medical instruments, furniture, toys, and computers has given us a broad view of the latest technologies available and has taught us how to do quality product development and how to do it quickly and efficiently.”

Learn something about everything

Learn as much as you can as fast as you can about the current state of the art surrounding any new project. The deeper and more flexible the knowledge you learn, the more easily you can use these past experiences to interpret new problems and recognize the value of past solutions.

“The best way to come up with ideas was first of all to go out and look at what’s out there.”

…knowledge brokers… may not know as much about any one thing as would a more specialized firm, but they often know when and where to look to learn more.

Find hidden connections

Build pathways that link project teams to the relevant knowledge of others in the firm. Invest in communication tools that provide for intensive interaction and analogic thinking. The more you communicate and interact around current problems, the more you will see these problems from a variety of perspectives and consider the range of possible solutions that this variety evokes.

Andersen’s Center for Strategic Technology demonstrates the firm’s latest thinking in business process solutions to its partners and clients. … not with brochures or technical specifications but, instead, with scenarios that combine a number of business and technical systems to paint a complex picture of what could be.

Make the idea work

Don’t stop with a good idea. Integrate innovative ideas with existing, well-developed, and well-accepted ideas from within the industry. Build prototypes, create simulations, and work with users to fit your innovative solutions into the established practices of these markets.

…implementation.. activities… generate a “wealth of knowledge that’s a result of the struggles, the agonizing they went through to try to figure out what’s the right way to proceed rather than the wrong way.”[2]

  1. “The Craic about “Fracking” – Technical Facts on Hydraulic Fracturing.”, 12 Aug. 2014, Accessed 20 Nov. 2016.
  2. Hargadon, Andrew B. “Firms as knowledge brokers: Lessons in pursuing continuous innovation.” California management review 40.3 (1998): 209-227.

Learning and communication drive projects

Learning by small teams and communication across projects drives productivity.[1]

Hiring and developing great people

“My first priority is to hire the best developers.”

Learning what customers want

The leading source of time delays in software development is rework: the redesign/recoding/retesting cycles made necessary by changes in requirements, changes in interfaces, etc.

To gain speed and productivity, managers must spend more time learning precisely what customers want in a software product and converting those wants into unambiguous specifications.

Devoting more resources to learning what customers want should be viewed as an investment, not just a cost.

…the high productivity firms tend to have a larger team devoted to determining customer requirements (which, if done right, may make coding more productive and less testing and correction necessary).

Front Loading is defined as the early involvement in upstream design activities of downstream functions– process engineering, manufacturing, and even customer service concerns.

Using proven components

When a component is reused in a subsequent product, the original design work is a form of virtual concurrency: in the initial effort, work is simultaneously carried out for all future products in which that component is used.

Concurrency across projects is the most difficult to visualize and accomplish, but also has the greatest downstream rewards.

Using small teams

…except for the customer requirements stage, faster firms tend to have smaller teams and smaller maximum team size. …smaller teams tend to be more productive (except in the early stages where input from many different sources is needed).

…adding bodies to a project to lower cycle time may have the opposite result because coordination and communication complexities make larger teams more difficult to manage.

…larger teams diminish productivity because of inefficiencies created by the difficulty of communicating within a large number of people. …communication demands increase in proportion to the square of the size of the team.

Helping information flow

…as you break problems into smaller pieces, interface complexities grow… …project managers… must ensure that the interfaces are simple and elegant.

…there are a number of important information flows…

Flying Start is preliminary information transfer flowing from upstream design activities to team members primarily concerned with downstream activities.

Two-way High Bandwidth Information Exchange is intensive and rich communication among teams while performing concurrent activities. The information flow includes communication about potential design solutions and about design changes to avoid infeasibilities and interface problems.

…project managers must establish two-way high bandwidth flows of information among the teams working on separate pieces of the problem…[2]

  1. Curtis, Bill, Herb Krasner, and Neil Iscoe. “A field study of the software design process for large systems.” Communications of the ACM 31.11 (1988): 1268-1287.
  2. Blackburn, Joseph D., Gary D. Scudder, and Luk N. Van Wassenhove. “Improving speed and productivity of software development: a global survey of software developers.” IEEE Transactions on Software Engineering 22.12 (1996): 875-885.