**1. Introduction**

The findings related to the role of financial development for economic growth is ambiguous in the literature.1 While an extensive set of studies have stressed the impact of financial development on economic growth to be positive (Bekaert et al. 2005; Christopoulos and Tsionas 2004; Arestis et al. 2001; Xu 2000; Levine et al. 2000; Levine and Zervos 1998), other studies find financial development's effect on growth to be insignificant (Ho and Saadaoui 2022; Demetriades and Hussein 1996). More recent investigations utilizing time series data exploring the impact of financial development on growth find the effect to be weak or even negative (Cevik and Rahmati 2020; Nwani and Bassey Orie 2016; Adeniyi et al. 2015; Samargandi et al. 2014; Quixina and Almeida 2014). Such findings are of interest, particularly because in the context of other development outcomes, financial development has been shown to be beneficial for poverty alleviation (Rewilak 2017), enhancing the effectiveness for aid-recipient countries (Nkusu and Sayek 2004), promoting foreign direct investment (Desbordes and Wei 2017) and generating comparative advantage for manufacturing economies (Beck 2002).

Given these findings, many studies have explored the determinants of financial development, including institutional determinants (Roe and Siegel 2011; Herger et al. 2008; Beck and Levine 2004; Galindo and Micco 2004; Rajan and Zingales 2003; Johnson et al. 2000), policy determinants (Chinn and Ito 2006; Boyd et al. 2001; Huybens and Smith 1999) and economic development-related determinants (Levine 1997, 2003, 2005; Jaffee and Levonian 2001). We add to this strand of literature by exploring the role of linguistic structures for financial development. Specifically, we explore a specific linguistic trait—future time

**Citation:** Caskey, Gregory W., and Nabamita Dutta. 2022. Financial Development and Language Structures. *Economies* 10: 313. https://doi.org/10.3390/ economies10120313

Academic Editor: Franklin G. Mixon

Received: 18 March 2022 Accepted: 4 November 2022 Published: 8 December 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

reference—and investigate if it explains differences in financial development across countries along with other factors.

In recent years, a number of studies have investigated the effects of linguistic traits on economic outcomes (Chen 2013; Galor et al. 2016; Mavisakalyan et al. 2018). One in particular, future time reference (FTR), is a linguistic trait indicating whether languages require an obligatory future tense marking. Languages such as English and French are *strong* FTR languages, as these languages require a dedicated form when referring to future events. In English, we state *she will go to New York tomorrow*. The same in French is stated as *Elle ira à New York demain*. Words like "will go" and "ira" represent a marked future tense. In contrast, languages such as German and Finnish are *weak* FTR languages, as the same grammatical form can be used for the future and the present. In German, for example, the same sentence can be stated as *Morgen geht sie nach New York.* The word "geht" indicating "goes" can be used for present and future tense.

We hypothesize that financial development is benefitted in countries where agents speak "weak" FTR languages relative to countries speaking "strong" FTR languages. In the context of FTR, the notion of "temporal displacement" (Mavisakalyan et al. 2018) suggests that dedicated used of grammar to indicate future events can make the future very distant for the individual. In the case of these strong FTR languages, the future can potentially appear discontinuous to the present relative to speakers of weak FTR languages.

As emphasized in the literature, the five key functions of the financial system include ex-ante production of information and allocating capital, generating effective corporate governance, mobilization of savings into investment, efficient risk management and minimizing cost in financial transactions (Dutta and Meierrieks 2021; World Bank 2012). We argue that FTR has effects on these outcomes, particularly through the channel of individual differences in future discounting. For example, individuals speaking strong FTR languages may be likely to care less about savings and financial intermediaries under such language structures, are less likely to be bothered less about easing the cost of acquiring capital as that can be put away for the future.<sup>2</sup> Likewise, well-functioning property rights are considered to be an essential part of financial development. The incentive for structuring such operational property rights might be lacking for agents speaking strong FTR languages. Since the future does not seem connected to the present, essential functions of the financial system—allocating resources effectively, protecting investors, sharing of information—can appear less rewarding. Thus, financial development is likely to suffer in countries speaking strong FTR languages. In contrast, agents speaking weak FTR languages are likely to discount the future less, as they cognitively associate the connection between the present and the future relative to strong FTR language speakers, and thus, are incentivized to generate an effective financial system. Under weak FTR language structures, agents are likely to place effort in effectively allocating capital towards saving and investment, create systems that protect investors and shareholders, and work towards minimizing the cost of acquiring information about financial decision making.

Our contribution in this paper is adding to the extensive strand of studies that have looked into the determinants of financial development and highlighting the role of language structures. Our results show that countries speaking weak FTR languages are likely to have between 10 and 25% percentage more financial development (depending on the specific measure of financial development used) relative to countries speaking strong FTR languages. Our results are robust to the inclusion of an array of controls, including political institutions. We check our results to alternate measures of financial development assessing its different characteristics—financial depth, size of the financial system, efficiency and extent of equity market activities.

Section 2 provides a brief background and literature review. Section 3 explains data and the sources. In Section 4, we describe the empirical methodology and benchmark results. Robustness analysis is described in Section 5, and Section 6 concludes.

#### **2. Literature Review**

The idea that linguistic traits can influence thought, and thus have effects on human behavior, has long been investigated. Based on the works of de Saussure (1916) and Wittgenstein (1922), the Sapir–Whorf hypothesis (SWH) stresses the idea that language can influence thought. Subsequent lines of research have explored this idea in various context. Though the hypothesis has been supported by many studies, the seminal works of Chomsky (1957) and Pinker (1994) have contested these findings, stressing that languages do not shape human cognition or ways of thinking. In subsequent decades, the Linguistic Relativity Hypothesis (LRH) was developed, which advances the Sapir–Whorf hypothesis by stating that both human cognition and behavior can be shaped by languages. While LRH was regarded as misguided by some linguists and cognitive scientists (Mavisakalyan et al. 2018), a substantial and ever-growing body of literature emerged starting in the 1990s testifying to the validity of the theory (Levinson and Wilkins 2006; Kay and Regier 2006; Boroditsky et al. 2003; Slobin 2003; Levinson 1996).

The idea that there's a direct influence of language on cognition and behavior is at the heart of the Linguistic Relativity Hypothesis (LRH). In a nutshell, the LRH states that the structure of one's language has a systematic influence on cognition and behavior, and as such, different languages represent the world in different ways by emphasizing different aspects of reality. As a result, speakers of a certain language may be more sensitive to various features of the world relative to speakers of another language. For example, studies in psychology like Harner (1981) show that for children speaking English, the use of future tense begins as early as age 3. English is considered to be a strong FTR language or a language requiring a dedicated future marker. Szagun (1978) also investigate differences in FTR for English (strong FTR language) and German (weak FTR language). He found no differences in future verb usage among children but did find such differences being reflected among adults.

As Mavisakalyan and Weber (2018) point out, the studies in economics on the effects of language on social outcomes differ from those studies in linguistics and psychology in a few significant ways. Whereas the former utilize much larger sample sizes and focus on the connection between language and broader economic and social outcomes, those studies in linguistics and psychology tend to use smaller sample sizes and focus on smaller, more specific cognitive effects. While studies like Licht et al. (2007) and Tabellini (2008) have considered linguistic structures as a source of exogenous variation in culture (Mavisakalyan and Weber 2018), more recent studies have considered linguistic traits as proxies for culture (Bhalotra et al. 2015; Santacreu-Vasut et al. 2014). Some studies have focused on investigating the effect of linguistic traits on various outcomes, explaining the association through the channels of both culture and cognition, but without distinguishing well between the two (Hicks et al. 2015; Santacreu-Vasut et al. 2013). Studies like Chen (2013) and Mavisakalyan (2015) claim that linguistic structures affect behavior and, thus, outcomes by directly altering individual's cognition. Studies such as these form the base of the new LRH literature.

As an example of the future marker, we can give the example of English where specific words like "will" or "is going to" has to be used to indicate " it will snow tomorrow". On the other hand, a language like German can imply the same thing by stating *Morgen schneien es* and not using grammar to indicate marked future events. Chen (2013) emphasizes that languages requiring grammar to indicate marked future events or strong future time reference (FTR) languages have speakers that are less future-oriented behavior which, in turn, lead inefficient outcomes.3 As examples of inefficient outcomes, Chen finds that individuals speaking strong FTR languages save less, have less wealth after retirement, smoke more, tend to be obese and engage in unsafe sex.4

Galor et al. (2016) mention that speakers of languages that do not have marked grammar use indicating future tense are likely to have long-term orientation. Long term orientation or the lack of it affects individual's discounting of future. A reduction in an agent's discount rates can be because of long term orientation. Speakers of strong FTR languages are likely to discount much more relative to speakers of weak FTR languages (Mavisakalyan et al. 2018). As such, FTR is relevant to the long tradition of studies on the human tendency to discount future costs and rewards (Frederick et al. 2002; Kirby and Herrnstein 1995; Solnick et al. 1980; Ramsey 1928).5

#### **3. Data**

#### *3.1. The Sample*

Our sample consists of an unbalanced panel of 100 developed and developing countries over the period 2001 to 2018. The panel is unbalanced because our variables of interest are not available for each country for every year. Overall, we have 844 observations for 100 countries, or approximately 8.44 observations for each country. We compile our data from the World Bank's Global Financial Development Database (GFDD) and World Development Indicators (WDI), as well as from Chen (2013). Below we describe our variables of interest in detail.
