Abstract
We present a field experiment investigating the mechanism by which community currencies enhance trust. Our question is the following: do I trust more when using a community currency because I am a trusting-type person or because I think that you are trustworthy? We call the former preference-based trust; while the latter is belief-based trust. We apply a modification of the standard trust game from the experimental economics literature to disentangle these mechanisms. Player A has to choose whether or not to trust player B, and player B can either reciprocate that trust or not. Our innovation is in experimentally separating the currency in which the game is played (effective currency), from the currency preferred by the participant (preferred currency). If the mechanism is preference-based, then preferred currency will determine trust more than effective; if it is belief-based, then the effective currency will be determinant. We find strong evidence of the preference-based mechanism of community currencies on trust, and only weak evidence of the belief-based mechanism.
Introduction
More than a decade after the global financial crisis, citizens and governments alike in developed countries are still questioning the nature of globalized money and the impact it has on our lives in the short and long terms. Community currencies (CCs) and other alternative economic instruments are increasingly attracting the attention of consumers skeptical of both the credit-money capitalist basis (Michel and Hudon, 2015) and the permanent economic growth imperative (Dittmer, 2013). This skepticism and the interest in CCs as an alternative means find their source in a sense of disconnection between economy and territory due to financial globalization (Ruzzene, 2015). The breakdown of local communities and the regression of local economies due to globalization and deindustrialization since the 1970s can be seen as a reduction in levels of social capital (Jackson, 2019); rebuilding this has stimulated a search for a diversity of socioeconomic institutions in general and of monetary and financial ones in particular. Of particular importance may be the effect that this deterioration of local economy has on trust. Research has shown that generalized trust has concurrently fallen in developed countries since the 1970s, and tied that breakdown to increasing income and wealth inequality (Gould and Hijzen, 2017). Indeed, although social capital itself is a nebulous and notoriously difficult concept to define both clearly and completely (Jackson, 2019), a high level of generalized trust and trustworthiness has been defined as one of its core components (e.g. Scrivens and Smith, 2013). While they do not solve the structural problems in the financial system, CCs are seen in this context as a tool for regenerating social capital to create sustainable economies through strong local relationships, and particularly for generating trust as a social phenomenon. This has resulted in a huge variety of CC initiatives on an international scale (Nishibe, 2018). However, trust at a social level is built from microfoundations of individuals deciding whether to trust or not. This individual-level trust can moreover have many different sources and forms (Dasgupta, 2005; Glaeser, 2001). In this article, we use a field experiment to investigate the class of trust generated by a particular CC – the Eusko, in the Basque region of France.
Trust can be defined as voluntary exposure to strategic risk: to trust others means to place your own outcome at least partially in their hands. 1 The decision of whether or not to trust can therefore be analysed in a manner analogous to any other decision under risk. Expected utility maximizers will rationally choose to trust if the probability-weighted utility of the outcomes – in this case, the results of trustworthy or untrustworthy responses to trust – is greater than the utility of the – untrusting – outside option. This implies that there are two separate components to the trusting decision. The first is clearly the perceived risk of betrayal on the part of the trustee. If one expects trust to be honoured, then the trust becomes much easier to rationalize (Algan, 2018). We will refer to trust that comes from a perception of low risk of betrayal as belief-based (or signaling) trust. However, for a given perceived risk of betrayal, people may differ in their basic disposition, or ‘readiness to trust’, understood as the inverse of risk aversion applied to the strategic domain. Those who are more trusting will have a higher ‘certainty equivalent’ to the risky trusting decision, and will therefore choose to trust for a wider range of outside options, than those who are dispositionally less trusting. Indeed, the idea of a trusting individual is probably first understood as being something like risk-loving in strategic risk. This would imply a willingness to trust even when the net expected return is negative. Algan (2018: 290), for instance, calls this ‘Moral Trust’. In line with the link to risk preferences, we will refer to it as preference-based (or expressive) trust.
The reader will observe that either of these classes of determinants of trust can work through many empirical mechanisms. For instance, in cases where reputation is easily established, belief-based trust will be easier to sustain. Preference-based trust may be affected by social proximity between the agents, which again can have family, ethnic, national or ideological roots among many others, and other causes can be easily imagined. However, the analogy to standard decision under risk suggests that these many ‘causes in fact’ reflect basically two ‘proximate causes’: preference- and belief-based trust. Our article attempts to disentangle the classes without digging as deep as the specific determinants. These classes, moreover, are reflected in the pronouns in the article’s title: In short, if I trust you more in situation X than in situation Y, it could be because of ‘me’ (preference-based, higher readiness to trust), or because of ‘you’ (belief-based, higher subjective probability for your trustworthiness). Our basic research question is: within the CC community, is it you or is it me who increases the trust? That is, do I as a CC user have lower ‘social risk aversion’, and therefore trust more than someone using the national currency for a given level of perceived risk, or do I simply believe that you as a CC user are more trustworthy, and trust more for that reason? Key to our design, we experimentally separate the choice of currency from the effective currency in which the experiment takes place. If participation in the CC induces a ‘trusting disposition,’ for instance through priming of community norms, then those choosing the CC should trust more regardless of the actual effective currency in which the experiment takes place. On the other hand, if trust in the CC is induced by identifying more trustworthy partners and therefore reducing the perceived risks, then regardless of the initial choice, trust should be greater when the experiment is implemented in the CC than when it is in the national currency. Therefore, we can test which component dominates by measuring which variable better predicts trust, the choice of currency or the effective currency of play. In what follows, we describe how the Eusko fits into the general typology of CCs and the hypotheses we seek to test, then explain our experimental design and present results.
CCs and Eusko
We find four types of CCs in the literature. From a historical prospective, the first generation of CCs is represented by Local Exchange Trading System (LETS), which appeared in the 1980s, while the second generation emerged with Time Banks at the end of the 1980s. These are club-membership-based currencies. While LETS use virtual currencies for trading goods and services, Time Banks use time-denominated currencies for exchanging personal services (Dittmer, 2013). The third generation is represented by Hour Currencies, printed currency that circulates locally without any legal tender backing. In the context of financial crisis and its consequences associated with the statement of an economic growth sustained by monetary institutions (Ruzzene, 2015), the fourth generation of CCs started spreading in France after 2010. This type differs from Hour Currencies mainly in that it is backed by the national currency (Euro) and secured to it by a fixed exchange rate (Blanc and Fare, 2018). This CC is referred to as convertible local currency (CLC) by Dittmer (2013). As a currency type scheme, Blanc (2011) classifies LETS and Time Banks as CCs while Hour Currencies and CLCs are complementary currencies. However, unlike commercial complementary currencies such as super market loyalty points, CCs have two particularities, the first is being created by civil society rather than private institutions and the second is having limited circulation space (Dittmer, 2013; Seyfang and Longhurst, 2013).
In France, around 30 CLC projects exist today, backed by the social and solidarity economy law of 2014. The most dynamic French CLC is the French Basque Country’s EUSKO (Blanc and Fare, 2018). Meaning ‘Basque’ in Basque language, the Eusko CLC today has more than 3200 consumer users, 820 professional users and over one million euro-equivalent Euskos in circulation. Launched in January 2013 and managed by the Euskal Moneta association, Eusko is a CLC with an exchange rate of 1 Eusko:1 Euro. Two forms of Eusko are used today, physical bills and an electronic form using a bank card. Although in exchange Eusko nominal prices must be the same as those denominated in Euros, professionals who have a membership at the association can convert Eusko to Euro at the exchange rate of 1 Eusko: 0.95 Euro. The 5% fees are used to cover the association’s costs and to donate to local associations voted for by consumer users (2% and 3% respectively). Acceptance of professionals’ membership is conditioned on fulfilling at least two challenges in two years: (1) the first challenge is environmental and is related to selling at least three local products (or employing three Eusko members, or sorting the waste of the company); (2) the second is a cultural challenge and is related to presenting the products using the Basque language alongside with French (euskalmoneta.org, 2017).
CCs and trust
We are far from the first to study the relationship between CCs and trust. In a study on a Japanese CC initiated by a government program in the Tokyo suburb of Ichikawa, Richey (2007) cites four factors of trust promotion through CCs: (1) face-to-face contact and the volunteering-based communal spirit (feeling better about my neighbours), (2) healthy town-based local patriotism (feeling better about my anonymous townsfolks), (3) credit bartering (feeling more trustworthiness in people), and (4) expanding networks through cooperation efforts of the movement members. In an evaluative overview of local currencies, Martignoni (2012) evaluated 11 local currencies (both euro-backed and service-backed in German speaking countries) based on four principles: purpose, trust, creation, and circulation. The principle of trust questioned whether the issuance of local currencies is based on low trust in people (thus the currency should be backed by a higher value money) or is based on high trust in people (then local currency issuance is equivalent to pure personal loans creation).
In a study on the social effects of CC in Spain (particularly LETS), Sanz (2016) finds that as an outcome of reciprocity and participative behaviours in CCs communities, users perceived benefits related to improving social capital components including trust. In addition, 79% of the users reported increased satisfaction coming in part from positive perception of trust relationships. According to a definite majority of the survey participants of Jacob et al. (2004), the American Ithaca HOURS increased their social capital in terms of trust relationships creation.
Richey (2007) cites three factors of trust creation independently of CC presence in the community. The first is the socioeconomic status, the second is the active social networking, and the third is the psychological satisfaction with one’s city and with life in general. We must keep in mind that the CCs which Richey studied are those government-based local currencies that recompense volunteering. In a case study on an Australian LETS in Tasmania, Reputation based on ratings from community members is found to be an important instrument for increasing generalized trust in a CC (Krabbe, 2015). The usage of CCs (a study on six Kenyan CCs) is found to have a positive impact on local trust as well as on increasing trust in the community (Ruddick, 2015). A sociological analysis of the Sardex (a euro-backed Italian CC in Sardenia) by Littera et al. (2017) confirmed the key role of trust in addition to the fact that the operation of the circuit (or community) was found to be reinforcing trust.
Thus, it appears that the goal of increasing local levels of trust is important to the development of CCs, and the evidence suggests that this is successful in that CC communities do indeed feature greater levels of trust and trustworthiness as part of their overall development effects. However, the observational evidence in the literature so far may confound different sources of this effect, suggesting a potential role for experimental methods. In this article, we propose a field experiment to distinguish between the belief-based trust and the preference-based trust. The difference between the two is not unrelated to arguments such as that in Ellingsen et al. (2012) regarding the effect of frames on behaviour: do they change people’s basic prosocial attitude, or rather the expectations about others’ behaviour, which in turn triggers changes in their own? Belief-based trust stems from the fact that trust is a rational response to an expectation of trustworthiness. Under this idea, CCs increase trust because (the trustors believe) they identify trading partners most likely to reciprocate it. Thus, the CC reduces the perceived risk of the trusting decision but may leave the risk preferences unchanged. The CC, in this case, is useful primarily for learning about the others (as it signals their trustworthiness); a decision-maker trusts more when interacting in a CC than in a national currency simply because she believes the person she is interacting with is more likely to be trustworthy.
Our suggestion for preference-based or expressive trust is linked to expression of identity. The expressive value of an act indicates its power to confirm and underline certain beliefs and values, making them more prominent in an agent’s mind and therefore more determinant of behaviour (Anderson and Pildes, 2000). A substantial literature in law and economics underlines the expressive value of law, by which statement of laws, by expressing normative positions identify social norms that then become binding for individual decision-makers (Cooter, 1998). This expressive power has been found in experiments to be enhanced by commitment to the expression, which is often operationalized through voting for the law in question (e.g. Tyran and Feld, 2005). A personal expression is required for the value to be enhanced. We propose the possibility that when people choose to use the CCs, they are in effect expressing a commitment to community principles that the CCs uphold, and this commitment then constrains their later behaviour. In particular, it makes them more ‘trusting’, reducing the level of social risk aversion they feel.
The idea that medium of exchange may have normative value is not uncommon. Participants in previous experiments have been found to link the medium of exchange to aspects of prosocial identity. In particular, people seem to feel that exchanges for cash express a lack of prosocial engagement. Blood donors in Lacetera and Macis (2010), for instance, were more willing to donate in exchange for in-kind vouchers than for an equivalent value of cash. This may support the idea that exchanges made using a CC, which attempts to combine elements of market and community activity, may generate different intrinsic motivations than those made in national currency. Unlike the belief-based mechanism (which is primarily about interpersonal signaling, with the trustworthy individual signaling reliability to potential trustors, and trustors’ main interest in learning about the trustee), this expressive mechanism is primarily aimed at transmitting information about oneself, to oneself. It represents a personal commitment to the principles of the CC, generating trust potentially regardless of the trustworthiness of the trade partner.
Experimental design and hypotheses
Incentivized decisions are foundational in experimental economics. When payment for participation is not linked to the decisions made, results may fall prey to a number of hypothetical biases. In the study of normatively charged subjects such as trust, biases such as that for social desirability – a tradeoff in the respondent’s mind between honest reporting and saying what one thinks the researcher wants to hear – may be particularly problematic (Fisher, 1993). In our study, all the participants were given monetary incentives in Euros or Euskos. Our experimental game was conducted with 144 subjects. All the subjects were inhabitants of the French Basque country, approached in the city of Bayonne in October 2017. The game was conducted individually for each participant, and each interaction lasted less than 15 minutes. The average payoff for the game was 4.20 € (or Euskos (E) as the exchange rate with the Euro is 1:1).
We use a binary trust game, (see Appendix 1) a simplified version of the investment game of Berg et al. (1995). This is illustrated in the Figure 1. In our basic game, A has to choose Don’t Trust or Trust. 2 In the former case, A receives three currency units, and the game ends. In the latter, B receives 10 currency units and must decide whether to return one unit to A or five. The manner in which the currency was determined is the main design innovation in our experiment, and is described next.
Experimental design.
The experimental interaction was based on two sets of envelopes. Currency Envelopes implemented the randomization of the effective currency of the game. Return Envelopes contained B-player decisions to resolve the payoffs. We first created the Return Envelopes in each currency by collecting the choices of 12 players in the B-role (trustee). We explained the game, said that B was playing the role of a trustee, and that they would be matched with an A player who had trusted. B therefore received 10 currency units in the chosen denomination. 3 Finally, B was given a choice to return either one unit of the currency, or five. Seven B subjects accepted to play in Eusko (the first 6 decided to send back 5 Euskos). Five subjects played in Euro (the first 4 decided to send back 5 €). These decisions composed the contents of the two Return Envelopes (one in Euro and one in Eusko). Therefore objectively speaking, the rate of reciprocity was higher in Eusko than in in Euros (about 85.7% vs. 80%). However, this was not communicated to the A-role players, as we wanted their ‘natural beliefs’ about the different trustworthiness of those choosing the two currencies to be unaffected.
Currency Envelopes each contained tickets that determined the effective currency of the game. The Euro-Envelope contained 60% Euro-denominated tickets, the Eusko-Envlope 60% Eusko-denominated tickets. We approached other individuals for A decisions. We again described the game, and then asked A to choose a Currency Envelope to determine the currency in which it would play out. We described the currency envelopes as containing ‘mostly Euro-denominated tickets’ or ‘mostly Eusko-denominated tickets’, but did not give participants any information about the proportions. This design aimed to make the choice a salient ‘vote’ or expression of support one way or the other, but to generate a substantial rate of participants choosing one currency but playing in the other, as this was key in the predictions. A was then paid 3 units of the withdrawn currency. A was asked next to choose whether to keep the three units, or to play out the game based on the choice of one B player in the appropriate Return Envelope. 4 A was aware that the B players had already been paid at this point, and so while B’s choice would determine A’s payoff, A could no longer influence what that particular B received. This has the advantage of removing altruism as a confound for trusting behaviour. A’s choice was based entirely on the perceived social risk of trust.
Once the game final outcome had been determined, we paid participants, and recorded their gender and the answer to a question about whether they considered themselves regular users of the Eusko, as well as the value of the last purchase they had made in Eusko.
What are the expectations from this design? Our primary predictions turn on the difference between belief- and preference-based trust.
Prediction 1: if belief-based (signaling) trust is prevalent, then trust behaviour will correlate with the currency in which the game is played.
Prediction 2: if preference-based (expressive) trust is prevalent, then trust behaviour will correlate with currency chosen by the participant.
Two aspects of the randomization of the actual currency that A plays are important. The fact that A has a higher chance of playing in the chosen currency preserves the relevance of the choice. An A-player who prefers one currency has a clear incentive to reveal that preference through choice. On the other hand, the fact that choice does not perfectly determine the actual currency allows us to differentiate in a controlled manner between the effect of choosing to play in Eusko (expression) and that of actually playing in the currency (belief). In addition to the social desirability bias, another problem with observational data on this topic is to do with selection: the choice to trade in one currency is hard to empirically disentangle from the actual trade. It is rare in practice to choose to interact in one currency, but end up exogenously interacting in another. In this experiment, the randomization means that we have observations of the preferred currency of play that are independent of those of the actual currency of play, whereas these are highly correlated in observational data.
One aspect of the design to notice is that while we vary chosen and actual currencies independently, the result is an asymmetry between the two categories of the choice. Participants who randomly select their preferred currency may well react differently from those who play in the same currency by choice. We introduce another set of hypotheses to investigate these more second-order effects through an interacted regression. Defining binary variables Choose and Play to indicate the Eusko for the appropriate variables, we estimate trust with the equation
This will allow us to calculate the average effect of each variable for each value of the other. Table 1 shows the terms of this equation that are operative in each case. Table 2 outlines the secondary tests we draw from it.
Estimating second-order trust rates.
Note: The trust rate in each experimental cell is estimated by the sum of the regression coefficients in that cell of the table; tests are calculated as the difference across cells. For example, we test the hypothesis that those who choose Euro Trust more when the effective currency is Eusko against the null that βE ≤ 0.
Tests from the regression equation.
Note: The trust rate in each experimental cell is estimated by the sum of the regression coefficients in that cell of the table; tests are calculated as the difference across cells. One-sided hypotheses express the relationships developed in Predictions 1 and 2.
The tests labelled DC-€and DC-E look at the difference in trust by choice, given play in the indicated currency. These are vertical comparisons of the columns in Table 1. DP-tests compare trust rates horizontally across rows of Table 1, corresponding to the difference in trust by effective currency, given a choice. These are more restricted versions of Predictions 2 and 1, respectively. The DX tests test the diagonals of Table 1. DX is the ecologically realistic’ comparison for cases in which there is a segregation by currency. Each individual type is trading in the preferred currency. The diagonal comparison shows how this confounds the two effects, but it is useful to check that we obtain the higher trust in CLC contexts from the literature.
Data and results
Our analysis is of N = 132 participants in the A role. The key variables of interest are whether they chose to play in Eusko instead of Euros (Choose), whether they actually played in Eusko (Play), and whether they decided to trust the B or not (Trust). We also asked participants whether or not they were regular users of the Eusko (User), and recorded their gender (coded as Male). 5 Table 3 gives the general picture of our variables.
Descriptive statistics.
Note: The column ‘Proportion who…’ shows the proportion of the sample satisfying the criterion in each row. N = 132.
Table 3 shows that although one-third of participants claim to use Eusko on a regular basis, 42% chose to play in that currency. This implies that some participants who are not users nevertheless chose to play the game in that currency. This is evocative, and we will discuss it further below. Since our design implied that choosing a currency gave subjects a 0.6 probability to play in it, the 42% rate of choosing to play in Eusko implies an expected play proportion of 0.48, quite similar to the 0.53 proportion observed. Therefore, the randomization of play appears to have worked well. We note that the rate of trust (86%) is quite high, overall. 6
Our primary predictions 1 and 2 concern the overall effect of our choice and play variables on trust. Table 4 shows the Trust rates for various categories of other variables. The p value for a Chi-square test with one degree of freedom is also reported. We see that those who chose to play in Eusko trusted nearly universally (96%), a significantly higher rate than those who chose to play in Euros (78%, p = 0.003). Regular users are also significantly more trusting than non-users. Gender does not predict trust in our experiment, and crucially, the currency the game was actually played in is also insignificant.
Primary hypotheses.
Note: Cells show the rate at which participants in each category chose to Trust, and the count in each cell. N = 132. Chi-square results shown for a null hypothesis of no association between the row variable and the Trust choice.
Therefore, Prediction 1 about belief-based trust is not supported in our data, while prediction 2 about expressive trust is supported. Participants’ choice to play in one currency or another predicted their trust well, regardless of the currency in which the game was actually played. The currency itself, however, had no significant impact on trust. Result: Expressive trust is more prevalent in this experiment than belief-based trust.
Second-order results.
Note: Trust rates (N) by experimental cell.
In this table, the vertical comparisons show that the higher trust rates from choosing Eusko carry over to actual play in both currencies. All 18 Eusko choosers trusted when playing in Euro, while only 33 of 44 (75%) Euro choosers trusted when playing in Euros. When actual play was in Eusko, 35 of 37 Eusko choosers trust (95%), while 27 of the 33 (82%) of Euro choosers trusted when playing in Eusko. On the other hand, the horizontal comparisons indicate that neither Eusko choosers nor Euro choosers were dramatically influenced by the actual currency. Eusko choosers trusted nearly universally in both currencies (18/18 vs. 35/37), and Euro choosers’ trust rates were also similar. Somewhat surprisingly, for both values of the chosen currency, trust levels were higher when the actual currency was not the same as the chosen one. 7 To add rigor to this analysis, we consider the regression described above. Rather than present the full regression results, we reproduce Table 4 from above as Table 6, and add test estimates and p values.
Results from regression tests.
Note: OLS regression estimates, with tests done on the linear combination of coefficients noted.
The story that emerges from this table is the following. There are two basic types of individual in the population. One type gets expressive value from choosing Eusko, and then trusts no matter what the actual currency of play is. The other type chooses Euro, and then trusts slightly more when the game is played in Eusko due to a belief-based mechanism. Overall, the preference-based trust is much stronger than belief-based, and results in significantly more trust in the Eusko environment than in the Euro environment.
We close this section with a brief discussion of the difference between Eusko users and non-users. As the reader might conjecture from Table 4, the fact that regular users and those choosing to play in Eusko have similar trust rates is in large part because they were the same people. Among the 44 regular users of Euskos, 35 (80%) chose to play the game in that currency. These percentages were reversed for non-users; only 23% of non-users (20 out of 88) chose to play the game in Eusko. An interesting question is: Which is the better predictor of trust? Logistic regressions of Trust first on the variables alone, and then together, find that both are significant alone, but that the coefficient is higher for choosing to play in Eusko than for being a regular user, and that when both are present, choice retains its significance while being a regular user does not. This is interesting in that it suggests that the act of choosing to play in the local currency by itself has an effect on behaviour, over and above that of selection into the ‘cooperative’ group that uses the currency regularly, and we take it as evidence of the expressive value of participating in the Eusko community.
Conclusion
Our experiment contributes to the literature by isolating the choice to engage in CC interactions from the belief about the choices of trading partners, in order to tease these components apart. The crux of our design attempts to disentangle trust based on belief from trust due to commitment to local currency ideals. If local money works by segregating the population into a cooperative subgroup that uses the money and an uncooperative group that doesn’t, then people who choose to play in Eusko, but end up playing in Euros, should know that the B with whom they are associated chose Euros, and thus is signaling an uncooperative type. Therefore, the hypothesis of belief-based trust is that it should be the currency of play that matters, not the choice. On the other hand, if choosing to play in Eusko represents a commitment to values of community that then frames or shapes the following behaviour, then the choice of currency should be a better predictor of play than the actual currency itself. In other words, the commitment hypothesis is that choice matters, not currency.
We find that the choice of trustors to use CCs is strongly predictive of trust regardless of the actual currency in which the game played, while the currency itself has no predictive power. Therefore, our results lean towards the expressive value of CC participation rather than the belief-based value. We see some weak evidence that those who choose Euros to begin with may later choose to trust more when the actual currency is Eusko, which is compatible with a belief-based mechanism, but this seems to be a minor determinant of the overall greater levels of trust in the Eusko than Euro environment. We also find that choice of currency is a better predictor of trust even than a self-report as a regular user. We would argue that this is consistent with the hypothesis that regular users are already committed to the social principles behind the local currency, and so they have less scope for increasing their trust than do non-users.
The distinction between the two trust mechanisms may be important for the development of CC systems. The belief-based mechanism of CCs would be useful in expanding trading networks to non-users who rationally prefer to interact with trustworthy types. However, it will be susceptible to opportunistic exploitation. Specifically, if untrustworthy types also begin using the CCs, then the beliefs in question will become unsustainable, causing them to unravel. Stated otherwise, belief-based trust may expand use of the CCs in the short run, but lead to problems related to the necessity for community in the longer term for the system. Evidence of the expressive mechanism of CC use, were it shown to exist, would be interesting in itself, but would also provide a barrier to this exploitation. If expressive use of the CC itself leads not just to trust but also to trustworthiness, then the unravelling will be limited. On the other hand, it may be that not everyone is susceptible to this kind of expressive modification of preferences; the community by nature may remain somewhat circumscribed. Therefore, the vision of the longer-term development of the CC system depends on the mechanism. Expressive trust will result in a relatively small-scale, but sustainable system, while belief-based trust may lead to greater growth but a less resilient model. We note that the former is much more in line with the principles of the CC movement.
We finish with a word about community and control. Based on the literature about the skepticism of CLC groups relative to global financial systems, it seems that the expressive value of engagement with the CC movement represents in part an attempt to regain control over an economic system that in globalizing, has passed beyond a human scale. The inherent limits to the scale of the CC, due to fact that it requires this engagement, is in this perspective not a drawback. It is, rather, representative of the fact that CLCs attempt to provide a real alternative model of economic development. We argued above that from a practical perspective the difference between the sources of trust was two-fold. On the one hand, expressive trust might limit the expansion of CLC programs relative to belief-based trust, because it relies on an underlying propensity to engage with the CLC ideals. Belief-based trust does not require anything from the trustor beyond the anticipation that trustees who play in Eusko are more likely to be trustworthy than those who play in Euros. On the other hand, a system that expands to include participants whose interest is purely strategic is by the same token susceptible to dilution and exploitation. However, these two components come together if the nature of ‘expressive preferences’ is such that even behaviour that begins strategically may lead in the end to the construction of trust in others, thus further develop the trusting norm in the CC community. 8 Our results therefore suggest that expressive trust mechanisms represent a positive track for the future development of CLC programs.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
