Bounded Rationality

Herbert A Simon was one of the most influential thinkers of the 20th century. Simon was born in Milwaukee, WI in 1916.  His father, Arthur, was an Electrical Engineer of Jewish origin who emigrated from Germany in 1903 and his mother, Edna, was a pianist who was a third generation American.

In 1975, Simon was awarded the Turing Award for his contributions to human cognition and artificial intelligence. In 1978, he followed that up by collecting the Nobel Prize in Economics for developing the theory of bounded rationality.

He was a polymath’s polymath. Simon graduated with a PhD in Political Science, did research in Psychology and Organizational Behavior, was a professor in Computer Science and Artificial Intelligence and won a Nobel in Economics. The unifying question that drove his life’s work over a 40 year span was to understand decision making and problem solving. His curiosity led him to pursue a doctorate, write dozens of papers and numerous books, and develop new theories that challenged our understanding of decision making. As of 2016, he is the most referenced author about artificial intelligence and cognitive psychology on Google Scholar [i].

Here are six things I learned Herbert A. Simon’s autobiography, Models of My Life.

1. Our ability to make rational decisions are bounded by the situation.
Simon was awarded the Nobel Prize in Economics in 1978 for his research on decision making within organizations. Much of his research focused on improving our understanding of how people make decisions. Traditional economic theory said that managers make rational, profit-maximizing decisions. What economists call rational choice. Always logical. Always prudent. Always utility maximizing.

Simon found that rational choice was an incomplete view. He described the firm as a complex, dynamic system. And within this system, people attempt to make rational decisions, but their capacity to do so is limited because of social connections, time constraints, and limited knowledge about the consequence of their decision. Put another way — the world is really complex, we don’t want to let others down and we have a lot going on — so we don’t fully think through our decisions. And, it’s pretty much impossible to consider all the information when making a decision and all the potential outcomes.

This leads people to make decisions that are satisficing — combining ‘satisfying’ and ‘sufficing’ — rather than maximizing.

This is what is known as bounded rationality [ii].

Bounded rationality states that people would not decide differently if they had more information or more time. Rather, it’s that people can’t process all the information even if it was available. Therefore if we want to anticipate people’s actions, knowing the amount or quality of information is not enough. Rather we should understand the cognitive process used to select from the information available. Hence people often seek solutions to decisions that are “good enough” using “rules of thumb”. These are known as heuristics.

As Simon notes:

We act out our lives within the mazes in which Nature and society place us.

Bounded rationality applies to individuals and organizations. Say you want to buy a TV, so you read a few online reviews and talk to a couple friends. When the sales person at the electronics store offers you a better deal on a comparable TV, you turn it down. This is because your reality has already been shaped by your friends and you are not willing to consider the other options.

In an organizational setting, let’s say a decision needs to be made. Rational decision making means that you will follow a logical, step-by-step approach analyzing the situation, considering facts, and make a decision that is made to produce the optimal outcome. However, the decision has to be made within a time constraint. Therefore bounded rationality is often manager – using a shorthand ‘heuristic’ to make the decision. It may be that your boss has made a similar decision or others won’t question the decision. It may not be the optimal decision for the company or the individual – but the manager will make a decision while not considering all the alternatives and select the first alternative that is satisfactory.

Finally, let’s say you are completing a survey. It’s been found that respondents choose answers that are satisfactory rather than optimal. This is because the respondents care more about what their answers signal to research more than they care about the accuracy of their response. This leads researchers to false conclusions.

A theory I would like to be explored is a rationality spectrum. Why doesn’t something like this exist? On one end, is complete rationality. On the other, bounded rationality. In each situation, there are certain variables which affect how rational the decision maker can be. Said another way, certain situations are more ‘bounded’ than others.

2. Our attention is finite. We must consciously filter information to make sound decisions.

Reference [iii]

Simon’s most famous quote is:

“a wealth of information creates a poverty of attention.”

The broader context is important to understand, his full comment was:

“In an information-rich world, the wealth of information means a dearth of something else: a scarcity of whatever it is that information consumes. What information consumes is rather obvious: it consumes the attention of its recipients. Hence a wealth of information creates a poverty of attention and a need to allocate that attention efficiently among the overabundance of information sources that might consume it.”

This is especially true in 2017. Technology brings more and more information to us in every conceivable moment. Simon developed the Model of Decision to help us filter information. The model has three steps: intelligence, design, and choice.

In the intelligence phase, the problem is identified as the decision maker ‘decides what to decide’. Problem searching involves comparing the actual state to the standard and problem formulation of the why. The output is a decision statement. The second phase is the design phase, which involves stating the objectives, research alternatives and analyzing the pros and cons.

In the choice phase, the alternatives are evaluated and the decision is made. Tools such as decision tree analysis are often implemented during this stage. The output is a decision to move forward with.

3. The General Problem Solver paved the way for artificial intelligence.
In 1939, when Simon was 33 he headed to Carnegie Institute of Technology, which would become Carnegie Mellon University.

He teamed up with Allen Newell and Cliff Shaw to develop a computer program called Logic Theorist in 1955, which was designed to imitate human problem solving. They were at the forefront of this field. In fact, the term ‘artificial intelligence’ was not coined until the following year.

The Logic Theory Machine was essentially the beta of the General Problem Solver. In 1957, Newell and Simon created the General Problem Solver (GPS) which had rules for solving problems based on the information processed.

GPS used means-end analysis as the central tool for problem solving. As Simon notes:

“…both Al and I, apparently independently, found in a particular thinking-aloud protocol clear evidence that means-ends analysis was the subject’s principal problem-solving tool. Means-ends analysis is accomplished by comparing the problem goal with the present situation, and noticing one or more differences between them—for example: I am here, I want to be there; I am five miles from my goal. The observed difference jogs memory for an action that might reduce or eliminate it (take a bike or an automobile; walk). The action is taken, a new situation is observed, and, if the goal has still not been reached, the whole process is repeated.”

4. We view the world differently because we employ different mental representations. These models encode information differently leading to different solutions for the same problem.

The representations one holds in their mind will shape how they approach problems. What representations do mathematicians use when researching problems? What about scientists, teachers, or artists? And where do these representations come from? While some think in words, others think in “mental pictures”.

Jacques Hadamard, the French mathematician, mentioned in his book, The Psychology of Invention in the Mathematical Field that mathematicians (specifically, American mathematicians) think in images. Hadamard noted in this letters to Einstein:

“[They] avoid not only the use of mental words but also, just as I do, the mental use of algebraic or any other precise signs; also as in my case, they use vague images.” [iii]

He followed up by saying:

“The words or the language, as they are written or spoken, do not seem to play any role in my mechanism of thought. The psychical entities which seem to serve as elements in thought are certain signs and more or less clear images which can be “voluntarily” reproduced or combined… The above mentioned elements are, in my case, of visual and some of muscular type.” [iv]

Simon believed that heuristic search was the right mental representation:

“One segment, under the banner “Let language lead the way,” takes verbal reasoning as its metaphor for the problem-solving process, and thinks of reasoning as some kind of theorem-proving structure. The second segment of the cognitive science community uses heuristic search through a problem space (a mental model of the task domain) as its metaphor for problem solving. Human Problem Solving (Newell and Simon 1972) adheres strictly to this viewpoint.”

5. The Cowles Commission created Econometrics and the General Equilibrium Theory and produced nine Nobel Laureates.

The Institute for Advanced Study in Princeton, NJ is one of the most notable independent research centers in the U.S., with over 33 Nobel Laureates who have passed through its doors including Einstein and Godel. While lesser known, the Cowles Commission was a formidable challenger. The Commission was formed at the University of Chicago in 1932 and had a twenty three year run till 1955. They had quite the run in 23 years — changing the direction of economics and producing 9 future Nobel Laureates.

How do they do it? Simon hints at the diversity and debate:

There was a mix of backgrounds and languages and strong accents spoken…The accents may have been more of a help than a hinderance to understanding. When several speakers tried to proceed simultaneously, by holding tight to the fact that you were trying to listen to, say, the Austrian accent, you could sometimes single it out from the Polish, Italian, Norwegian, Ukrainian, Greek, Dutch, or middle American. As impressive as the cacophony was the intellectual level of the discussion, and most impressive of all was the fact that everyone, in the midst of the sharpest disagreements, remained warm friends.”

The charter of the Commission was to link economic theory to statistics and mathematics [v]. Basically they wanted to utilize math proofs to validate economic theories. They held weekly seminars where a bunch of future Nobels got in a room and debated. It must have worked. The group is credited with creating two fields: General Equilibrium theory and Econometrics, as well as advancing linear programming, identifiability, and the simplex method.

During those twenty three years, the Commission output and influence on the field of economics is profound. Here are a few:

• Tjalling C. Koopman: Nobel 1975 for his theory of optimal allocation of resources.

• Ragnar Frisch: Nobel 1969 for founding econometrics.

• Kenneth Arrow: Nobel 1972 for his contribution to General equilibrium theory, also Arrow’s Impossibility Theorem. Also, he is the uncle of Larry Summers.

• Lawrence Klein: Nobel 1980 for his statistical models for economics.

• Leonid Hurwicz: Nobel 2007 for his mechanism design.

• Don Patinkin: monetary economics and money demand.

• Gerard Debreu: Nobel 1983 for his contributions to General equilibrium theory

• George Stigler: Nobel 1982 for his theory on the effects of public regulation on markets.

• Andreas Papandreou: became Prime Minister of Greece.

• Milton Friedman: Nobel 1976 for his consumption analysis and monetary theory

• Oskar Lange: market pricing in a socialist system.

• Trygve Haavelmo: Nobel 1989 for his contributions to econometrics and probability theory.

• Herbert A. Simon: Nobel 1978 for his theory on bounded rationality.

6. Simon’s thinking was influenced by farmers, authors, economists, political scientist and engineers
Simon’s work spanned multiple industries. His ability to combine art and science led him to develop theories and insights that many had passed over. Below are the people and experiences that had an outsized impact on Simon’s world view:

• Arthur Simon: Simon mentions the following about engineers, and his father by extension as he was an electrical engineer, “engineers believe in real things like machines and bridges and land. They are less confident that intangibles like money and organizations really exist, and the Great Depression enforced their skepticism.” This is part of what shaped his early beliefs.

• Harold Guetzkow: college classmate, friend, and shaped Simon’s thinking for over 25 years

• Maurice Davis: While Simon traveled the world, he says the real adventure of his life was Rockmarsh, a cattle farm 40 miles northwest of Milwaukee that covered nearly 3 square miles. Davis was the founder of the farm who was a manic depressive from his experience in World War I. Simon lived with Davis and two other young men on the farm. Simon credits that down-home experience outside academia as the single most important experience to develop his interpersonal skills. The farm eventually failed – from Davis’ suicide, to a widespread fire and a bout of pinkeye amongst the cattle – but this failure shaped was a wakeup call for what is controllable for Simon: “while any theories, however plausible and valid, can be destroyed totally by the obstinate facts of the real world.”

• Dick Cyert & Jim March: published the The Behavioral Theory of the Firm in 1963, which informed Simon’s belief that the decision-making process within the firm was essentially a problem-solving process.

• Jacob Marschak: head of Cowles Commission from 1943-1948, known as the founder of Econometrics, introduced independence axiom, and laid the groundwork for portfolio theory

• Allen Newell: computer scientist and cognitive psychologist at Carnegie Mellon and the RAND Institute. Worked with Simon to develop Logic Theorist and General Problem Solver.

• Yuji Ijiri: Cco-authored the Skew Distributions and the Sizes of Business Firms paper.

• Tjalling Koopmans: while Simon was teaching at the Illinois Institute of Technology he started attending the Cowles Commission seminars. Koopmans was one of the leaders who organized these sessions, which Simon said were his ‘second education in economics’.

References:
[i] Herbert A. Simon, Wikipedia https://en.wikipedia.org/wiki/Herbert_A._Simon
[ii] Simon coined this term in 1947 in his book, Models of Man, Social and Rational- Mathematical Essays on Rational Human Behavior in a Social Setting
[iii] MrunaltPatel. Herbert A. Simon, Wikipedia: https://en.wikipedia.org/wiki/Herbert_A._Simon#/media/File:Simons_3_stages_in_Decision_Making.gif
[iii] Jacques Hadamard, The Mathematician’s Mind: The Psychology of Invention in the Mathematical Field. Princeton University Press, Princeton NJ. 1945. Pages cited: 84, 143, 104-106.
[iv] Albert Einstein, Ideas and Opinions, Broadway Books; Reprint edition (June 6, 1995).
[v] The History of Economic Thought. http://www.hetwebsite.net/het/schools/cowles.htm