Chapter III: The Limits of Economic Modelling.

Although economic modelling is an indispensable tool for analysing and predicting economic
behaviour as discussed in Chapter I, its unrealistic assumptions and oversimplifications raise
important limitations. The emergence of behavioural economics and empirical studies makes it
possible to re-evaluate these models and offer more realistic perspectives. For effective public
decision-making, it is crucial to complement theoretical models with empirical data and
interdisciplinary approaches.
This chapter explores these limitations by drawing on key critiques and illustrations from the
literature and behavioral economics.

III.1 The unrealistic nature of the assumptions

III.1.a Homo-economicus !

Homo-economicus is a fundamental assumption of classical and neoclassical economic theory,

representing a simplified and idealized view of human behavior. This concept is based on the ideas of
perfect rationality, complete knowledge, individualism, temporal consistency, and risk aversion.
However, this view is largely challenged by research in behavioral economics.

1) Limited rationality : Herbert Simon (1955), in A Behavioral Model of Rational Choice,

introduces the concept of limited rationality to describe human decisions in contexts where
individuals are confronted with cognitive, informational, and temporal constraints. Rather than
looking for the optimal solution to a problem, individuals adopt so-called "satisfactory" strategies,
i.e. solutions that are considered good enough for their situation.

We can see the characteristics of limited rationality in:

- In the search for a satisfactory solution, individuals do not explore all possible options, but stop at the
first solution deemed acceptable. This reduces cognitive load and allows for faster decision-making,
Example: Rather than comparing all the products available on the market, a consumer can choose a
brand they know or a recommended option. This "good enough" approach reduces the time and
effort required to make a decision.
- Incomplete knowledge, individuals do not have access to all information or are not able to process it
optimally due to cognitive limitations.
- Using heuristics, rather than comprehensively assessing the consequences of each choice, decision-
makers use mental shortcuts to simplify their analysis. Example: A company may adopt a strategy that
seems to work in the immediate future, even if it is not optimal in the long term, due to the
complexity of economic forecasts or organizational constraints.
 The concept of bounded rationality challenges the assumptions of classical models based on strict
maximization. He emphasizes the importance of designing more realistic models that take into
account human limitations. For example, behavioral economics relies on these principles to explain
behaviors such as procrastination, decision-making under uncertainty, or time inconsistency.

2) Cognitive and emotional biases : Kahneman and Tversky (1979), in their seminal study on
perspective theory, showed that individuals do not make decisions on a strictly rational basis. On
the contrary, they are influenced by systematic biases:
- Loss aversion, individuals value losses more than equivalent gains. For example, the pain of losing
€100 is perceived as more intense than the pleasure of winning €100.
- A framing effect, the way an option is presented can modify the choices. For example, a medical
intervention with a 90% chance of success is perceived differently than an intervention with a 10%
chance of failure, although the two are identical.
- Confirmation bias, individuals tend to seek information that confirms their pre-existing beliefs while
ignoring information that contradicts them.
- Heuristic of availability, decisions are influenced by the information that is most easily accessible in
memory, often emotionally marked or recently experienced.
These biases, whether cognitive or emotional, lead to behaviors that diverge from classical rational
predictions. They also question the relevance of homo-economicus as a model of human decision-
making.

3) Social preferences: Contrary to the homo-economicus model, individuals often take into account
the consequences of their decisions on others. These social preferences include altruism,
reciprocity, and aversion to inequality, which significantly influence economic choices.
- Altruism, individuals are willing to sacrifice part of their own utility to increase that of others. For
example, donations to charities or community mutual aid illustrate this behaviour.
- Reciprocity, economic choices can be motivated by a desire to reward actions perceived as
benevolent or to punish those deemed unjust, even if this entails a personal cost.
- Inequality aversion, individuals prefer balanced outcomes, even if it means reducing their own gain.
Illustration with the prisoner's dilemma
The prisoner's dilemma, a classic framework in game theory, illustrates how social preferences
influence decisions. In its classic version, two prisoners must choose between cooperating or
denouncing the other. The rational solution (according to the homo-economicus model) is to
denounce, because this is the optimal choice regardless of the behavior of the other. However,
 experiences show that many participants choose to cooperate, even if it entails personal risk. This
trend can be explained by:
Reciprocity : Participants cooperate in the hope that their partner will do the same.
Inequality aversion : Individuals avoid an outcome where one of the two suffers a sanction while the
other benefits.

4) Social norms : Cooperation is often perceived as a morally valued behaviour, which motivates
participants to act altruistically.
- These observations show that economic models need to incorporate social and emotional factors to
better reflect actual behaviors. The prisoner's dilemma highlights the limits of perfect rationality
assumptions and the need to take social preferences into account in economic models.
5) Temporal inconsistency: Classical economic models assume that individuals' preferences are
constant over time. However, behavioral studies show that these preferences can change across
time horizons, a phenomenon called temporal inconsistency. This means that individuals tend to
overestimate immediate benefits over long-term ones, even if the latter are larger, these can be
illustrated in the two concepts of procrastination and present bias:

Procrastination: Individuals often delay actions that they know will be beneficial in the long run. For
example, a student may choose to watch a TV series rather than revise for an exam, even though they
recognize that studying would be better for their future.

Present bias: This bias reflects a disproportionate preference for immediate rewards at the expense
of future benefits. For example, in a study by Thaler (1981), participants preferred to receive $50
immediately rather than $100 in six months, although the rational choice would have been to wait.

In a famous study, Mischel's (1972) marshmallow test experiment tested self-control in children. They
had the choice of eating a marshmallow right away or waiting 15 minutes to get two. Many chose
immediate gratification, illustrating the present bias. This experience reveals that short-term choices
are often out of step with long-term goals.

III.1.b Incomplete or asymmetric information

The perfect information hypothesis, central to models of pure and perfect competition, is rarely
encountered in reality. In many markets, economic agents do not have the same information, which
leads to information asymmetry. This type of situation can have a negative impact on the efficiency of
markets and their performance.

1) Disconnection from the reality of the markets: Cournot's and Bertrand's models
The models of Cournot (1838) and Bertrand (1883) illustrate how imperfect competition moves away
from the idealized assumptions of pure and perfect competition. Unlike the latter, which assumes a
large number of suppliers and buyers with homogeneous products, these models highlight strategic
behaviors specific to oligopolistic markets.

Cournot's model: Companies simultaneously decide on the quantities to be produced, taking into
account the quantities produced by their competitors. This strategic interaction results in a different
balance than that predicted by perfect competition, with higher prices and lower production. For
example, in an oil market where a few large companies dominate, each player adjusts its production
according to the strategies of the others, which distorts the results predicted by classical theoretical
models.

Bertrand's model: In this model, companies set their prices rather than their quantities. When
products are homogeneous, intense competition pushes prices to be equal to marginal cost.
However, in the presence of product differentiation, firms may maintain higher prices, such as in
technology goods markets where each product has unique characteristics.

These models show that, in reality, strategic interactions and market structure strongly influence
economic outcomes, thus moving away from predictions based on perfect competition. They stress
the importance of considering the particularities of real markets in order to build more accurate
economic models.

2) Adverse Selection and Barriers to Entry: Information asymmetry can manifest itself in a variety
of ways, but it has significant effects on adverse selection and barriers to entry in markets.
George Akerlof (1970), in his seminal paper The Market for "Lemons": Quality Uncertainty and
the Market Mechanism, highlighted how problems of uncertainty about product quality can
disrupt markets.

Adverse selection: Akerlof illustrates the phenomenon of adverse selection through the used car
market. Buyers who do not have complete information about the quality of the vehicles assume that
all cars have average quality. This perception discourages sellers of good quality cars (also known as
"peaches") from participating in the market, which essentially leaves "lemons" (low-quality cars) for
sale. The result is a gradual collapse in the average quality of products and, potentially, in the market
itself. This same mechanism can be seen in other industries, such as health insurance, where
insurance companies can set high premiums to compensate for unknown risks, thus excluding healthy
individuals from the market. This reduces the efficiency of the system and increases costs for the
remaining members.
Barriers to entry: Information asymmetry also creates barriers to entry into markets. For example, in
the pharmaceutical industry, new companies find it difficult to compete with existing players due to
the lack of comprehensive information on drug research and approval processes. This dynamic
reinforces the dominant position of large companies and limits competition.

III.1.c ceteris paribus! and closed systems

The ceteris paribus hypothesis, meaning "all other things being equal," is a crucial simplification used
in economic modeling to isolate the effect of one variable while assuming that all others remain
constant. Although this tool is useful for theoretical analysis, it is often criticized for being out of step
with reality.

1) Limitations of the hypothesis : In real economic systems, variables interact in complex and
dynamic ways. For example, higher fuel taxes can not only reduce gasoline consumption, but also
affect production costs, employment levels in energy-related sectors, and even environmental
policy. Ignoring these interdependencies can lead to erroneous or incomplete conclusions.
2) Closed systems: Traditional business models often assume closed systems, where external
interactions are neglected. However, in a globalised economy, exogenous factors such as financial
crises, pandemics or climate change play a decisive role in the development of markets. For
example, the COVID-19 pandemic has highlighted how external shocks can disrupt global supply
chains, something that ceteris paribus-based models cannot capture.

III.2 Possible consequences of oversimplification

III.2.a Over-generalization
One of the major criticisms levelled at economic modelling is the risk of over-generalisation.
Theoretical models, by simplifying complex phenomena, can produce erroneous conclusions when
applied indiscriminately.
1) Risk of erroneous conclusions: Robert Lucas (1976), in his famous article Econometric Policy
Evaluation: A Critique, points out that traditional econometric models often fail to take into
account changes in the behaviour of economic agents in relation to public policies. This
phenomenon, known as the "Lucas critique," highlights the inability of these models to predict
structural adjustments to the economic system in response to external interventions. For
example, the 2008 financial crisis demonstrated the limitations of traditional macroeconomic
models, which had underestimated the impact of speculative behaviour and interconnectedness
in the banking system.
This over-generalization of economic dynamics has led to inaccurate forecasts and inadequate
policies in the face of the crisis.
 2) Underestimation of structural factors: Traditional economic models tend to neglect structural
factors such as social inequalities, political contexts or institutional structures. However, these
elements play a decisive role in economic dynamics.
- Concentration of wealth and economic growth, Thomas Piketty, in Capital in the Twenty-First Century
(2014), shows that the excessive accumulation of wealth in the hands of an elite slows down long-
term economic growth. By analyzing historical data over several centuries, Piketty reveals that when
the rate of return on capital exceeds the rate of economic growth, inequality widens, threatening
social and economic stability.
- As practical examples, models that ignore these factors often underestimate the barriers to access to
resources for the poor, limiting their active participation in the economy. For example, unequal
education systems reduce economic opportunities, leading to long-term productivity losses.
- Importance of institutional structures, political and legal institutions shape economic incentives. For
example, inefficient judicial systems or unstable regulatory frameworks can discourage investment
and impede economic growth.
III.2.b Poorly calibrated public policies
Poorly calibrated public policies often come from simplified economic models that ignore the
complexity of human behavior and market dynamics. The homo-economicus hypothesis neglects
cognitive biases, such as loss aversion or the control effect, thus reducing the effectiveness of the
measures. For example, a carbon tax can fail if citizens perceive the immediate costs as too high
compared to the long-term benefits.
In addition, failure to take into account social preferences, such as altruism or aversion to inequality,
can lead to social resistance, especially in sensitive reforms such as pensions. Tensions increase when
policies ignore the collective dimensions, hindering their implementation.
Perfect information assumptions also omit actual market asymmetries, leading to imbalances.
Without regulation, adverse selection phenomena, such as in the health insurance market, or barriers
to entry encourage monopolies, limiting innovation and widening inequalities.
Finally, policies based on ceteris paribus underestimate side effects and external shocks. A carbon tax
without consideration for its impact on employment can amplify inequalities. Moreover, crises, such
as the one in 2008, reveal the shortcomings of simplified models, which are unable to predict and
respond effectively to systemic collapses
Examples:
- Nudges, such as default choices, have been shown to increase organ donation rates (Johnson &
Goldstein, 2003). However, universal application of this policy without local adaptation is likely to fail
due to cultural differences.
- Public guarantees or certifications in markets such as organic farming are necessary to restore
consumer confidence and compensate for information asymmetries.
- Policies to reduce plastic bags (Homonoff, 2018) have been successful thanks to the integration of
behavioural biases, such as loss aversion. A purely theoretical approach would have ignored these
dynamics, thus limiting their effectiveness.

References

o A Behavioral Model of Rational Choice - Herbert Simon (1955)

o Theory of Perspectives - Kahneman and Tversky (1979)
o The Market for "Lemons": Quality Uncertainty and the Market Mechanism - George
Akerlof (1970)
o Capital in the Twenty-First Century - Thomas Piketty (2014)
o Econometric Policy Evaluation: A Critique - Robert Lucas (1976)
o Nudges and Default Choices - Johnson and Goldstein (2003)
o Reduction of plastic bags - Homonoff (2018)
o The models of Cournot (1838) and Bertrand (1883)