Language and Economics

Language and Economics

language.jpgThe FINANCIAL -- In the 1930s, the American linguist Benjamin Lee Whorf put forward the hypothesis that people of different mother tongues perceive the world differently.


According to linguistic relativity or Whorfianism, both the grammatical structure and the vocabulary of a language influence the way how people think. Proponents of political correctness, aiming to ban the usage of certain words that are considered to be derogatory or discriminatory, ultimately base their ideas on Whorfianism.


Saying “little person” instead of “midget” may have an impact on how one thinks about people who have the medical condition of dwarfism.   

These ideas were picked up by other groups as well. The Khmer Rouge in Cambodia prohibited the usage of the personal pronoun “I”. In this way, they wanted to force people to see themselves as part of a collective rather than as individuals. There are also instances of Whorfianism in fictional literature. In his famous novel "1984", George Orwell imagined a society in which the government banned subversive words, making the associated thoughts unthinkable.    

Assuming that Whorf had a point, would it not to be expected that language also influences the economic behavior of people?


Modern economic theory is very much concerned with the role of information in economic systems. In this vein, the famous Israeli game theorist Ariel Rubinstein wrote a whole book on economics and language. Also empirical economics is catching up. In a recent study, Yale economist Keith Chen claims that the language used by a nation has an impact on saving behavior and health precautions (“The Effect of Language on Economic Behavior: Evidence from Savings Rates, Health Behaviors, and Retirement Assets”, American Economic Review, 103 (2013): 690-731). 

Chen argues that the way in which present and future tenses are formed in a language influences a speaker’s intertemporal choices.  Some languages force their speakers to indicate that they are explicitly talking about something that will happen the future, while other languages allow people to talk about the future essentially in the same way as about present events. According to Chen, languages that grammatically equate future and present (i.e. the formation of present and future tenses does not differ much) foster future-oriented behavior. Apparently, people who speak such languages save more, retire wealthier, smoke less frequently, and are less often obese.

“If language nudges you to speak about the present and the future in the same way, is it possible that you feel about the present and the future in the same way?” asks Chen.

The article divides languages into two broad categories: weak and strong future-time reference (FTR) languages. For instance, English is a strong FTR language, requiring to talk about the present and the future differently. We can say "it is cold today", but if we're referring to tomorrow's weather, we have to say "it will be cold tomorrow." Not all languages make this distinction. For example, in the Finnish language one would say "today be cold" and "tomorrow be cold", and one can also say "be cold" without specifying whether one is talking about the present or the future. Therefore, Finnish is categorized to be a weak FTR language. Mandarin has no tenses at all – a sentence in Mandarin usually does not convey information on whether the speaker refers to the present or the future. This does not mean that Mandarin speakers are not able to understand time difference; however, they are not forced to select a tense each and every time they make a statement. So, also Mandarin is a weak FTR language.

For measuring the degree of future-orientation of different nations, he looks at their spending habits, their retirement savings, and their health precautions. Then he relates these values to the FTR type of the language according to the Typology of Languages in Europe (EUROTYP) classification. His analysis includes 76 developed and developing nations (including Georgia) on five continents. Based on an econometric regression, Chen provides support for the hypothesis that speakers of weak FTR languages are more future-oriented, both in monetary and non-monetary matters. Compared to people who speak strong FTR languages, those whose mother tongue is weak FTR on average have saved 39% more capital by the time they retire, and the probability to have saved a positive amount in a given year goes up by 31%. If one is a weak FTR speaker, the probability to be a smoker goes down by 24%, to be obese by 13%, and the probability to be physically active increases by 29%.

Chen claims that this partly explains why the Unites States, Greece, Israel, the United Kingdom, Poland, Turkey, Italy, and other strong FTR language countries on average save less compared to Norway, Switzerland, Japan, Netherlands, Finland, Austria, Belgium, and other weak FTR language countries.



According to the study, Georgian is a strong FTR language. Does this explain the abysmal saving rate of Georgians and the alleged lack of “safety thinking” that was discussed on the ISET Economist blog two years ago?

Georgian is a very special language. The Kartvelian language family has a grammatical structure that differs significantly from those of the Indo-European, Semitic, and Asian languages. The distinct features of the Georgian language might be reflected in the national character of Georgians. As this topic has not been picked up yet by psycholinguists and behavioral economists, let us engage in some free speculations.

Georgians distinguish between elder and younger persons by addressing them “შენ” (you) “თქვენ”(you). Perhaps, this formal expression of respect (or juvenility) creates subconscious borders. Intuitively, it induces a hierarchy among people that would not be there without the language. Whether that makes cooperation easier or more difficult is unclear though. In some teams and organizations with a flat hierarchy structure, the Georgian language may create artificial obstacles reducing the effectiveness of the group. In other settings, however, hierarchy may be useful. Languages like Japanese and Korean (the latter has 18 different forms for addressing different people) may enable their speakers to more easily integrate into given structures and execute tasks more reliably.  

In another blog post, published about half a year ago by Norberto Pignatti and myself, we analyze the gender pay gap in Georgia. Some of those who commented on the article started a fierce discussion on whether our analysis was right or not, and some claimed that the gender pay gap was not related to discrimination in any way. In support of this viewpoint, they did not refer to the Georgian language, which they could have done, because Georgian may indeed be less “sexist” than other languages.

Georgian does not have a grammatical gender and not even 3rd person pronouns. Georgian personal and demonstrative pronouns yield only deictic (contextual) and number information, yet they do not indicate the gender of the person referred to. Intuition suggests that this should reduce the discriminatory potential of the Georgian language. In their language, Georgians remove any difference between men and women, they are all the same.

Other languages, like German, attribute a gender to each and every object. A table is male, while a curtain is female. The floor is male and the wall is female. The psychologist Thomas Jacobsen has found that these grammatical genders have an influence on how Germans perceive different objects (“Effects of Grammatical Gender on Picture and Word Naming: Evidence from German”, Journal of Psycholinguistic Research, 28 (1999): 499-514). Lacking grammatical gender, there can be no such thing in Georgian.

Moreover, in German the word “man” often has the meaning of “somebody” (even if that person is female). To avoid this sexism, politically correct people in Germany do not say “man” anymore, but replace it by “human”. The Georgian language does not have this problem: even the staunchest Georgian sexist has no choice but to use politically correct language.