In this brief literature review, I have attempted to summarise two distinct viewpoints of nature of preferences. I start with outlining literature on how preferences are created or where do they come from; traditional approach of assuming that preferences are stable and evidence from behavioural economists disproving that and concluding that preferences are dynamic and reversible in nature. In the last section, I present a case of coarse grains in India and conclude with the objective of this study.
How a consumer reacts when her income and prices of goods and services change, is one of the most interesting aspects of economics. The simplest explanation is Marshall’s law of demand which states that demand curve is a downward sloping curve meaning that as price decreases, the demand for the product will increase assuming all other variables (such as income) stays constant. The downward sloping curve is derived from diminishing marginal utility which suggests that as consumption of a good increases, after a point of time its marginal utility goes down with every additional unit consumed.
Thus, in order for the consumer to purchase more and more units of the said good, price must decrease to attain more or equal level of utility. It is safe to say here that Marshall more or less ignored the income variable in determining demand of a commodity especially by assuming that marginal utility of money remains constant thus putting the income variable out of the utility function. In the words of J.R.Hicks (Value and Capital, 1939),
“…Marshall’s law states that if his [consumer] income increases, and the price of X remains constant, the price of X will still equal the marginal rate of substitution, without any change in the amount of X bought. The demand for X is therefore independent of income. His demand for any commodity is independent of his income…”
In the same chapter, Hicks introduces the idea of the income consumption curve or the expenditure curve. It is an upward sloping curve that relates consumption and income given that the prices remain unchanged. With the help of indifference curves, Hicks shows that consumption or expenditure of an individual will increase when her income increases and vice versa. This is true for all “normal” goods. An exception to this theory is “inferior” goods. Inferior goods according to Hicks are those goods whose consumption is fully or partially replaced by normal goods as the consumer’s income increases.
In other words, the income effect here is negative such that at higher levels of income, expenditure on inferior goods would decrease assuming unchanged prices. Common examples of inferior goods are coarse grains, public transport, margarine, off brand goods, generic medicines, coarse clothes (made from fabrics considered inferior such as jute) etc. However it is important to note here that what is considered inferior and what is not is highly subjective. A person concerned with environmental costs of fuel would not find public transport inferior or people from under developed and developing countries might not find generic medicines inferior. Fast food for example, can be considered inferior by some too. Hence, inferior or normal is a very subjective preference.
My research is going to talk about preferences a lot, especially their nature. What is a preference? A google search of this word will show that preferences in simple terms refer to a liking, bias or a penchant for a choice over other options available. In both Economics and Psychology, preferences are always assumed to exist and are really important determinants of decision making. All consumer theories for example Cardinalist approach (Marshall), Indifference curve and Ordinalist approach (Hicks), Revealed Preference theory (Samuelson) to name a few, are based on the assumption that consumer’s choice indicates her underlying preferences that are utilized when she makes a choice. How preferences are made is a subject matter that is extensively debated.
There are two major approaches that attempt to reason the existence of preferences. First is the traditional economics approach that suggests that preferences already exist from past experiences and choices. Consumption patterns over time are made by changing and adapting preferences based on previous decisions. It is this learning that forms consumer preferences and affects the next choices (West et al., 1996). Second is the constructive approach that says that preferences are constructed based on the context and choices available (Griffin, Slovic & Tversky, 1990; Johnson, Payne & Brettman, 1993).
What information is available at the time of making a decision is a significant variable that affects constructing of preferences. Various theories based on this include endowment effect (Medvec, Madey, & Gilovich, 1995; Loewenstein & Issacharoff, 1994; Camerer, 1992) contingent valuations (Kahneman, 1993), Asymmetric effect (Simonson & Tversky, 1992) and others. Traditional economists however disagree with this approach by reasoning that it is impossible that an individual is able to completely forgo her previous experience and treats each decision as a unique one and that over time preferences attain stability at least in particular domains (West, 1996; Hammond, Stewart, Brehmer, & Steinmann. 1975; McClelland, & Mumpower, 1980).
Fischhoff (1991) in his paper on preferences concludes that while constructive approach hold true in the case of new and different products, after the first purchase a consumer does gain an experience of the product and develops preferences either favorable or unfavorable. After this point, the consumer’s behavior can be aptly described by the economic approach.
It is clear from above discussion that both approaches on their own are incomplete and it is their combination that tells us how preferences are made. Hoeffler & Ariely (1995) argue that it is true that preferences exist and consumers are able to break them down in attributes that they like. The example they use are baby strollers.
A new parent wants the stroller to be strong and durable as she relates these attributes to safety. Hence, while she has never purchased strollers before, she has a preference towards the attributes she wants the stroller to have. They further discuss how learnings from previous experiences and decisions happen and their pattern over time to conclude that with experience, the preference for certain attributes will get stronger and stronger and will stabilize over time. George Hommans (1961), in his revolutionary paper on human decision making and sociology, first introduced the Rational choice, which was extended by Cook (1977) and Blau (1964) into a more methodical and formal theory.
They define the rational choice theory as an explanation of how people make decisions – the individuals here are seen as affected and motivated by their preferences which they then make into a decision given certain constrains. The constraints here include prices and income among others. As it is impossible to have any function without constraints, individuals compare and relate the various choices available [in a purely technical form] by predicting the outcomes of each of the available choices and then choosing the one with highest utility.
Heath (1976) writes that individuals are rational and they make those choices or choose options that have higher chances of giving them satisfaction. Rational choice theory is very dynamic in the sense that it provides a strong base to analyze how rationality comes into play, whether or not it is affected by more information, certainty or uncertainty, self-interest or interest of others and any other relevant motivations. All traditional economic theories are based on the rationality assumption where the decision maker (DM) makes logical and well-judged decisions that provide her with highest utility given constraints.
Preferences according to rational choice theory are stable and exogenous. In Stigler and Becker’s De Gustibus Non Est Disputandum (1977), they use tastes and preferences as a single term. They claim that tastes and preferences are not fickle and neither do they differ significantly among people. More or less, tastes are stable and similar. They give a very interesting analogy to explain this by comparing preferences with mountains that are there and will continue to remain there year after year and look the same to all men.
So then, what determines consumer behavior one might ask. Their answer is prices and income. Their hypothesis asserted that “assuming tastes and preferences remain unchanged”, an important assumption in economic theories, should be completely removed as it held no meaning because preferences are stable over time as well as similar and that there exist utility maximizing functions that are stable and well-behaved. To support their hypothesis, they present their arguments against the research on existence of dynamic preferences.
In Marshall’s Law of demand, he does not account for time to be permitted for any change in existing preferences. He accepts their dynamic nature by giving the example of music where its exposure results in higher demand of it in the sense it becomes addictive. Stigler and Becker do not agree with this and instead call this addiction as “consumption capital” of the consumer. Likes and dislikes can be explained by the amount of time spent on the concerned activity and that preferences have nothing to do with it. If one spends time on addictive substances such as alcohol and drugs and give it time frequently to form into a habit, is another example for consumption capital.
J.S.Mill in his “A system of logic (1972)” gave the example of a king and how his decisions and policies are nowhere motivated by his personal interest. Rather, all policy decisions are largely influenced by feelings, sentiments, habits and mentality of his community members. This suggests that a consensus is possible and there exists similarity in the preferences of people.
Stigler and becker studied three other instances – advertisement, fashion and habitual behavior and provided alternative explanations suggesting that they have nothing to do with preferences and can be explained otherwise. Fashion or style is an example of ‘whats new’ and in the same way is circular. People want to constantly choose that which is new, there is some sort of public admiration to be gained from it and social rivalry in identifying the next new thing. However, fashion too is more influenced by prices and income rather than anything else.
Advertisement, popularly an important influencer of tastes (Galbrait, 1958) is also provided an alternate explanation. Stigler and Becker argue that advertisement provide consumers with relevant information that they might not previously have for example a soda drink having vitamin C or other such properties. Here knowledge is the important factor but eventually it is the prices and income that determine the demand for the product/service.
In the early 80’s, many farmers, agricultural economists and concerned sections started observing a steady decline in the consumption of red meat (beef) in United States and Australia (Christensen & Manser,1976; Pope, Green & Eales, 1980; Martin & Porter, 1985; Chalfant, 1987; Thurman, 1987).
The dominant theory was that there had been a structural change in the tastes and preferences of the combined population such that they preferred white meat (chicken) over red. This was mainly thought because of dietary concerns as beef contained high levels of saturated fat and increased cholesterol levels. Such reasoning that it is the changed preferences that had caused significant decline in the demand of red meat led to meat industries start expensive advertisement campaigns and large scale promotion tactics to reverse the preferences again.
Chalfant and Alston (1988) in their paper – Accounting for Changes in Tastes, tested whether shifts in demand was due to change in tastes and in doing so they also tested for stability of preferences, an important axiom of the revealed preference theory. To do this, they used time series data of per capita meat consumption of Australia and United States where they studies 5 different types of meat – lamb, mutton, chicken, pork and beef during 1962-1984.
Statistics showed that there had been an increase in chicken meat consumption and a definite decline in beef consumption. However, when a comparison was made between relative prices, Chalfant and Alston (1988) concluded that relative prices of red meat had increased and that of chicken meat had decreased.
To further check their hypothesis, they divided the different types of meat into bundles. If a bundle having more beef and less chicken was preferred earlier but later was rejected over a bundle with more chicken and less beef assuming constant prices, then preferences were said to have changed. Instead they found that the initial bundle was rejected over the later because it had become more affordable. This clearly pointed out that it was the change in prices and expenditures that had led to the demand shift and that there was no shift in preferences. Similar results were found by Landsburg (1981) who studied meat consumption shifts and patterns in the United Kingdom; Varian (1982); Swofford and Whitney (1986) among others.
Behavioral economics, a branch that came into existence only a few decades ago, uses concepts from psychology such as personality traits, cognitive theories, attitudes and motivation to explain economic phenomena especially concerning consumer behavior. The discipline strongly opposes the rationality assumption and the preference theory suggesting stable preferences. Many behavioral scientists have proven the inconsistency of this assumption and have found numerous instances where transitivity theorem doesn’t hold and the idea that the decision maker chooses that alternative which he predicts would give him the highest utility and that there exists such a utility function.
To give an example, let us consider an incentivized laboratory experiment (Slovic & Lichtenstein, 1968), participants are shown two lotteries – lottery A and lottery B where she is asked to choose one over other. The tradeoffs of both the lotteries are informed to her. Lottery A is a high probability lottery while the other is maximum winning lottery. Lottery A refers to a situation where a random dart is to be thrown on a circular board with a line. If the dart hit the line, the participant would receive zero payoffs and if it is hit anywhere else on the circle, she would receive four dollars. This is a lottery with high chances of winning money.
On the other lottery B had a situation where only if the dart manages to hit interior part of the circle will the participant receive sixteen dollars or else nothing. B, while giving maximum rewards had a low probability of winning. When asked how much the price of each lottery should be, majority placed a higher bet on B. When asked which lottery would the participants prefer? Surprisingly, majority preferred A over B. This violates the axiom that suggests that preferences are stable and transitive.
The argument here is that if individuals are able to construct utility functions of all alternatives available and that the choose the one with optimum utility, they wouldn’t place a high bet on B and at the same time choose A. Such a result is inconsistent with preference theory. Several researchers have proved inconsistencies in axioms of rationality theories such as completeness, transitivity and independence. They include Camerer (1995), Rabin (1998), Allias (1953), Ellsberg (1961), Bernoulli (1738) and Kahneman and Tversky (1992) to name a few.
Deutsch & Gerard (1955) suggested that social influences are significant variables that affect psychological processes concerning consumption behavior and that they also lead to preference reversal phenomena. They also go one step further to differentiate between various types of social influences such as normative and informational where the former refers to conformity to positive expectations of others. Here “others” is a subjective term that can include self, another individual or member of a collective group.
Normative social influence mainly talks about the real and perceived expectations towards self. Informational influences, on the other hand refer to influences by alleged evidence and proofs of reality. Thus, in their paper they explore the effects of social expectations and information on self-esteem, self-approval, feelings of anxiety and guilt – all of which leads to explain the various choices and decisions made by the individual. They conclude that preferences are subject to social influences and that they can be reversed. The experimental method they used was a modified version of Asch (1951) where the subjects are shown 3 lines.
They are then given an additional line and are asked to determine which initial line it comes closest to. Such scenarios are presented where accurate judgments are really easy to make. Their choices are to be made publically. Participants are divided into groups where except for one, all are informed by the instructor to deliberately make wrong judgments, this is done to see whether the uninformed individual will make a choice based on own judgment or due to influence of others. Similar conclusions are found in Cialdini & Goldstein (2004); Kahneman & Tversky (1982).
There are also theories that suggest that individuals have some amount of knowledge about the patterns of their preferences and that they are able to predict preference changes. Strotz (1955) introduced Sophisticated behavior approach which states that individuals are aware of the influence of present decisions and choices on their future preferences and based on this information, they make rational choices.
Yanoff and Hansson (2009) underline a unique dilemma in accepting the preference reversal phenomena. On one hand it is possible to explain any and every behavior based theory by based on the assumption that individual’s tastes and preferences are subject to changes over time and on the other, we’d again have to go back to rely on “attitudinal and introspective” approaches.
From the above literature review of preferences, it is clear that there is still no consensus in the field of Economics regarding stability of preferences. While Behavioural economists have proved that there are inconsistencies in human behaviour and that preferences are subject to change, traditional economists still assume their stability.
