Difference in the response to business gifts: former communist Poland versus the Netherlands Milou Ottink 10676651
University of Amsterdam Faculty of Economics and Business
MSc Business Economics: Organisation Economics 15 ECTS
Supervisor: Thomas Buser
Second examiner: Jeroen van de Ven Date: 29-06-2015
Abstract
It has been shown that gifts are able to influence the behavior of the receiver in many different situations and presents are widely used within businesses. However, there is a lack of research on cross-cultural comparisons concerning the perceiving of business gifts. This experiment
fills in this gap by comparing post-communist Poland and the Netherlands in how they perceive the receiving of a business gift. An experiment is conducted using the vignette method by providing the participants with information about a business negotiation involving
a gift. They are then asked questions related to how they feel about the deal and gift. After this, extra information about the deal is provided, framing it either as a zero-sum or win-win situation. Then the questions related to how they feel about the deal and gift are asked again. This design allows expectations about the participants’ way of thinking to be tested. Contrary
to any expectations, the results show that Poles respond more negatively to a business gift after receiving additional information framing the business negotiation as a win-win situation.
After receiving additional information framing the deal in a zero-sum way, both nationalities respond negatively with Dutch answer more negatively than the Poles as expected. However,
this difference between nationalities is not significantly strong. Moreover, there does not appear to be a difference in the strength of the responses between large and small gifts.
2 1. Introduction
It has been widely acknowledged and tested that gift exchange can have substantial effects for the behavior of the receiver (e.g Falk, 2007; Beltramini, 1992; 2000). This effect has been incorporated in the business world. Beltramini (1992) wanted to learn more about the effect of business gifts on attitudes and behavior in real life businesses by performing a controlled field experiment. This showed that a small business gift increased how positive the customer thought about the attributes of the product. That businesses use gifts for their expected positive effects can be seen by the fact that there is a large amount of money spend on business gifts (Brennan et al., 2006). Business gifts ranging from small to large are common in a variety of industries. This large volume and use increases the importance to learn more about the subtleties that are present in gift giving.
There are several proposed explanations to the use and effectiveness of gift exchange, from which reciprocity is the main one. Reciprocity means that giving something, such as a present, can make the receiver feel obliged to do something back (Cialdini, 1985). However, it is important to take the perceived intentions of the gift giver into account when dealing with reciprocity. It has been shown that people will not just reciprocate based on the consequences of a certain action, but also take the intentions that are behind this action into account (e.g. Falk & Fischbacher, 2006; Falk, Fehr, Fischbacher, 2008; Greenberg & Frisch, 1972; Lowe & Goldstein, 1979). Moreover, the size of the gift can also lead to different responses. In general it is important to take into account that there is both an outcome effect in the form of the size of the gift and an intention effect as in what someone tries to achieve by giving the gift (Falk & Fischbacher, 2006). If intentions are perceived to be good or bad a larger gift will thus give stronger results by strengthening the outcome effect.
It is important to realize that intentions can be perceived differently when looking at different cultures. This makes it valuable to understand more about the differences that affect business negotiations between post-communist countries and the West. If cross-cultural business gift giving is not done in the right way, this could deteriorate potential new business relations or stop fruitful negotiations (Arunthanes, Tansuhai & Lemak, 1994). This leads to the following research question:
3 This research contributes to previously investigated effect of gift giving, since experiments testing the effect and causes of gift giving are mostly conducted in western countries. Besides this there are to my knowledge no empirical cross-cultural comparisons in gift giving. This gives a gap in the literature and this study attempts to fill this gap by looking at the cross-cultural differences in the perceived gains and losses after receiving a business gift. In this study Poland and the Netherlands are compared to find out if there are cultural differences that may lead to a gift being experienced differently. Poland is an interesting country to look at, since it is one of the healthiest and fastest growing post-communist countries in the European Union. It is a high income economy, because of its booming market economy (Rountree, O’Kreefe & Chadraba, 2014). This makes it an attractive country to do business with for many Western countries including The Netherlands. The Netherlands and Poland have a high trading volume and the Netherlands is actually one of Poland’s main investors. However, it should be realized that Poland’s post-communist past gives it a totally different set of cultural values than the Netherlands has (Kolman, Noorderhaven, Hofstede & Dienes 2002). Poles do not necessary think in win-win situations as the Dutch do. Their mentality is more: if you win, than I must be losing (‘’Onderneem Polen’’, 2012). This can have
substantial effects on how a business gift is received. Dutch will see it as a nice extra, while Poles might believe they paid too much if the other party still has money left to give them a gift (‘’Onderneem Polen’’, 2012). This means that business gifts that are given as
appreciation of past negotiations can be experienced in a negative way in Poland, while Dutch can see it as something positive.
In this study Poland and The Netherlands are compared in their responses to receiving a hypothetical gift in a vignette study. This can help to show a difference between the countries that is more specific than the broad cultural contrasts that are shown to exist (Kolman et al., 2002). The insights that this study can give are an important contribution, since they can lead to knowledge that can be used in real-life negotiations such that businessmen trading with Poland know what to pay attention to. Moreover, it can lead suggestions for further research in this area, for example the extension of potential results to other countries with different cultural backgrounds.
The findings of this experiment show that Polish participants always think more negatively after receiving additional information about a business negotiation involving a business gift. This is the case even when the information frames the negotiation as a win-win situation and
4 that is against any expectations. When the additional information states the negotiation more negatively in a zero-sum way, both nationalities respond to this by seeing the deal and gift more negative as expected. Moreover, the Dutch appear to be responding stronger in their negativity than the Polish participants, which is what was hypothesized. However, the difference between nationalities is not very significant. Furthermore, looking at small and large gifts separately does not reveal a clear indication of large gifts giving stronger results than small gifts.
First some related literature will be addressed. Second, the methodology of the study and the hypotheses will be presented. Third, the results will be shown. Fourth, the results will be discussed and a conclusion will be given.
5 2. Related literature
2.1 Gift giving in general
Gifts can have substantial effects for the behavior of the receiver, this has been widely acknowledged and investigated (e.g Falk, 2007; Beltramini, 1992; 2000). Falk (2007)
conducted a field experiment on gift exchange to look at the consequences of a gift in a natural setting. They worked together with a charitable organization to send over 10,000 letters to households. There were three treatments, one without a gift, one with a small gift, and one with a large gift. The results proved the importance of the gift, since it was found that donations to a charitable organization can increase from 17 percent after a small gift to as much as 75 percent after a large gift (Falk, 2007).
Gift exchange is also extensively used in the workplace. As explained by Akerlof (1982) workers can provide a ‘’gift’’ by working harder than the minimum standard and the firm can give a gift in the form of a wage that is higher than the wage that the worker could get in another company. Kube and Maréchal (2012) continue on this finding by performing a controlled field experiment to test how both a monetary gift (20 percent wage increase) and a non-monetary gift (thermos bottle of the same value) influence a workers’ performance. They let workers catalog books in a library and the results show that a non-monetary gift has a strong effect on a workers’ productivity by increasing it with 25 percent. While the monetary gift shows no significant difference in productivity. The explanation is that in certain
situations non-monetary gifts can influence behavior more than monetary gifts, because ‘’it is the thought that counts’’.
2.2 Gift giving in business
The fact that gifts can have substantial effects has been incorporated in the business world. Beltramini (1992) was one of the first to recognize the lack of research on the impact that the extensively used business gift can have. To learn more about the effect of business gifts on attitudes and behavior in real life business situations, a controlled field experiment was performed. In this experiment there were several randomly assigned groups of customers, from which some received a business gift (leather business card case valued at $10.00) and others did not. Afterwards, the customers received a mail questionnaire in which they were asked to fill in their attitude towards several product attributes: price, quality, service, and
6 delivery. The results then showed that hat the small business gift made the customers’
perceptions of product attributes more positive (Beltramini, 1992).
The established positive effects that a gift can have are widely used within business’ strategies and can be seen by the large amount of money spend on business gifts. Extreme examples can be found in the pharmaceutical industry, where it is estimated that this $250 billion industry spends an astonishing $19 billion per year on marketing, largely in the form of gifts to US doctors. The estimations say this means that the spending runs up to $18,000-29,000 per doctor per year (Brennan et al., 2006). Even though the pharmaceutical industry is an extreme example, business gifts ranging from small to large are common across most industries.
Malmendier and Schmidt (2012) incorporate the effect that a gift is often at the expense of a third party in their study. In the case of business gifts we can see an example in lobbying. Suppliers try to influence procurement managers for favorable treatment by gift giving. This can influence the manager to disregard other suppliers, who might have better or cheaper products. There is a negative externality of gift giving here, since it comes at the expense of the shareholders of the procurement manager, who now do not have the optimal supplier. Malmendier and Schmidt (2012) use a controlled laboratory experiment to test the effect of gifts which have a negative externality, by coming at the expense of a third party. The experiment is designed so that a decision maker has to choose between two products, which are represented by simple 50/50 lotteries. The decision maker is instructed to make a decision that is best for his client and he gets paid a fixed wage that is not influenced by his choice. However, the decision maker can get a gift from an interested party before he makes his choice. This gift is given before any payoffs are known to prevent any informational content. Also, it is given no matter what the choice will turn out to be, the choice is completely anonymous and the participants are re-matched per round. These aspects make sure the experiment is designed in such a way that there are no monetary rewards for favoring the gift giver. Furthermore, in the main treatment none of the participants (both client and producers) find out the decision that was eventually made. Even though the experiment is designed in such a way that it would be predicted that the decision maker would not favor the gift giver over the client, the results show that the effect of the gift is stronger when it is at the expense of a third party. This is important to realize, since there are many occasions where business gifts involve a third party.
7 2.3 Reciprocity
The underlying causes of gift exchange have been examined in recent years with several proposed explanations. Reciprocity is the main explanation, which according to Cialdini (1985) means that giving something, such as a gift, can make the receiver feel obliged to do something back (positive reciprocity), but is also means that when something unkind is done, this will be punished (negative reciprocity). It turns out that even when people deal with complete strangers in a situation where reciprocating is costly and there is no
chance of any present or future material rewards, people tend to repay gifts and take revenge on hostile actions (Fehr & Gächter, 2000).
There is an abundance of both field and lab experiments which show the effect that
reciprocity can have. Several field experiments show the applicability that reciprocity has in real life. Tidd and Lochard (1987) show positive reciprocity in a field experiment where a waitress that gives a broad smile gets more tips than with a minimal smile. The field
experiment of Falk (2007) discussed earlier is also a good example of how reciprocity affects behavior, since it is shown that gifts increase the donations to a charity by a significant amount. Moreover, large gifts increase the donations by 75 percent, while small gifts only increased donations with 17 percent. These higher donations happened even though the letters with the gifts were send no matter what the amount of donations turned out to be. Another interesting example of reciprocity is the fact that after people have been given a free sample, they often tend to feel obliged to reciprocate this by buying something (Cialdini, 1993). Negative reciprocity has also been shown to exist in field experiments. For example, Kube, Maréchal and Puppe (2006) performed a labor market field experiment to investigate (among other things) the effect of cutting wages on work morale and effects. This was done by hiring students to catalogue the books of a library during 6 hours. At the beginning the students were told they would earn 15 euros per hour, but in the negative reciprocity treatment the students were told shortly before they started working that the wage would be only 10 euros per hour. The results of the experiment show that the negative reciprocity seems to be strong, since the reduction in wage has significant and lasting negative effect on the workers’ efforts (Kube, Maréchal & Puppe, 2006).
If reciprocity is tested in field experiments it is difficult to make it absolutely sure that the participants do not expect any future material benefit from reciprocating. This makes it necessary to also look at evidence for reciprocity in the controlled environment of a
8 laboratory experiment. In laboratory experiments it can be determined that the participants will face a real and often relatively high monetary costs of reciprocating, while it can be made sure that there is no future material benefit from the reciprocal behavior. Positive reciprocity in lab experiments is mostly tested by the use of trust games and gift exchange games. In a trust game it is the case that a proposer has an amount of money, from which he can send an amount between zero and all to a responder. The amount of money send is then multiplied and the responder can send between zero and all back to the receiver. Berg, Dickhaut and McCabe (1995) use a trust game in which the amount of money send (part of their $10 show-up fee) is tripled. The experimental design ensures complete anonymity and the game is only played once. This rules out any chance of reputational concerns related to repeated
interaction, contractual pre-commitments or potential punishments. The results of this lab experiment show that positive reciprocity is at work, since generally a positive amount of money is send and a positive amount of money is returned (Berg, Dickhaut & McCabe, 1995). A gift exchange lab experiment also involves a proposer and a responder. Here the proposer can offer a wage and decide if he makes the wage higher than minimum or not. After this the responder decides whether to accept or reject the wage. If the wage is rejected both get a payoff of zero and if the wage is accepted the responder decides how much (costly) effort to provide and if he provides extra effort or not (Fehr & Fischbacher, 2002). Without reciprocity it would be predicted that the equilibrium is that the responder offers minimal effort and that any positive wage is accepted. However, several lab experiments show that there is a positive correlation between the effort chosen and the wage offered (e.g. Fehr, Kirchsteiger & Riedl, 1993; Charness & Haruvy, 2002). This indicates that the responder will reciprocate high wage offers with higher effort choices (Fehr & Fischbacher, 2002).
Negative reciprocity is mostly tested in laboratory experiments by the use of the ultimatum game. The ultimatum game goes as follows. There is a proposer and a responder, the proposer is provided with a certain amount of money and can then propose a division of this amount between him and the responder. The responder then decides to accept or reject the proposal, if he rejects both get nothing. If both the proposer and responder would be rational and selfish, the expectation is that any positive amount of money send it accepted. The proposer would then send an amount equal to the smallest possible number and the responder would accept (Fehr & Fischbacher, 2002). However, laboratory experiments show that offers below 20 percent are rejected with a high probability and that this probability decreases with the size of
9 the offer (Camerer & Thaler, 1995). A low offer is thus negatively reciprocated by the
responder with the rejection of the offer (Fehr & Gächter, 2000; Fehr & Fischbacher, 2002). Early research on reciprocity found that the amount of reciprocity returned is equal to the amount of benefaction given, with the general idea that one should ‘’do exactly unto others as they have done unto you’’ (Pruitt, 1968; Wilkie & Lanzetta, 1970). Nevertheless, more research on the topic of reciprocity showed that receivers actually return more than what is given. This was confirmed in a laboratory experiment conducted by Regan (1971). In this experiment the results indicated that the normative pressure the reciprocate was the cause of the participants buying more expensive raffle tickets from a person who gave them a small gift in the form of a soft drink. This shows that reciprocity can cause small gift to have larger effects. The labor market field experiment of Kube, Maréchal and Puppe (2006) in which the participants were asked to catalogue library books had both the negative reciprocity treatment discussed earlier and a positive reciprocity treatment. In the negative reciprocity treatment the wages were cut before starting, while in the positive reciprocity treatment the wages were unexpectedly increased. It was found that there is an asymmetry present in the field between positive and negative reciprocity, since the results in the negative reciprocity treatment were much stronger (Kube, Maréchal & Puppe, 2006). This indicates that reciprocal responses can be even more important to take into account if something is perceived negatively.
2.4 Intentions in reciprocity
In general it is the case the people like to help others that helped them and feel like hurting the ones who hurt them (Rabin, 1993). When looking at the effects of reciprocity this makes it important to take the perceived intentions of the gift giver into consideration. People will not just reciprocate based on the consequences of a certain action, but also take the intentions that are behind this action into account (Falk & Fischbacher, 2006). A good
example comes from Ong (2011), who explains an event at Yale that showed that when a gift was insinuated as a possible bribe half of the people returned the gift the next day.
Furthermore, there are several studies which show the importance of intentions in reciprocity. Even though some studies show mixed results depending on which game is used, several experiments indicate that intentions do play an important role and should definitely be taken into account when reciprocal behavior (both positive and negative) is in place (e.g. Falk, Fehr, Fischbacher, 2008; Greenberg & Frisch, 1972; Lowe & Goldstein, 1979). The study of Falk,
10 Fehr and Fischbacher (2008) finds that when playing the ultimatum game, the rate of rejection depends on the offers that are available to propose. If there was a more equal option available and an unequal one is proposed, rejection rates are much higher than if this equal option was not available, but only a more unequal one was. This confirms the idea that negative
reciprocity is not fully determined by the final material payoffs, but that the intensions matter. The experiment was conducted with the use of four different mini-ultimatum games with different options to propose (Falk, Fehr & Fischbacher, 2008). However, it is a limitation that each of the subjects participated in all four ultimatum games and could thus have been
influenced by previous games played or be aware of what was tested. The study of Greenberg and Frisch (1972) also confirms the importance of the intentions within reciprocity. Their empirical study was conducted to be able to show the relationship between intentionality and the likelihood of reciprocation. In the experiment the participants were businessmen and could receive either deliberate of accidental help and either a high or a low amount. The results showed that when giving help was deliberate instead of accidental, the amount of motivation and obligation to help back was greater. This finding was explained in the
experiment by DePaulo, Brittingham and Kaiser (1983) as the difference between appropriate and inappropriate help. In their experiment subjects had to work on several problems and could earn appropriate help (help on problems for which help was necessary) or inappropriate help (only help on problems for which no help was needed). Appropriate helpers were
thought of more positively than inappropriate helpers (DePaulo, Brittingham & Kaiser, 1983). The indication of these findings is that in the case of case of business gifts it is important that the receiver feels as if the gift is appropriate and intended just for him (Beltramini, 1992). Another interesting result is that in a chance game experiment with monetary stakes from Gergen, Ellsworth, Maslach and Seipel (1975) recipients prefer the subjects who donated money and asked for something in return to the ones asking for nothing (Gergen, Ellsworth, Maslach & Seipel, 1975). This means that people can prefer to have a negotiation with someone who will ask for something back, if this is done in the right way. This suggests that after giving an appropriate business gift, it can actually be good to ask for something, such as future business, to restore the equilibrium in the relationship (Beltramini, 1992). However, this needs to be done with great care for how the gift is perceived. When receivers feel that there are ulterior motives this can also make them less reciprocal (Lowe & Goldstein, 1979), this makes gift giving a business that needs to be done really tactfully. A negative point in the studies mentioned above is that the subjects were manipulated in the experiment. They were told that the experiment was about something different than it actually was or they were made
11 to believe that they were playing against another person, when they actually were not
(Greenberg & Frisch, 1972; DePaulo, Brittingham & Kaiser, 1983; Gergen, Ellsworth, Maslach & Seipel, 1975).
2.5 Other proposed explanations
Besides reciprocity there are several other proposed explanations of the responses to gift giving. It could actually be argued that it has nothing to do with social preferences, but be purely because of standard economic theory. As explained by Schmidt and Malmendier (2012) a repeated relationship or informational content can cause the responses to a gift to be perfectly rational. They give the example of a physician who chooses to prescribe more drugs to a certain pharmaceutical company after the attendance of a conference sponsored by the company. This can be the rational thing to do for the physician if he thinks it can give him more sponsoring in the future or if he received information on the conference that taught him that this was the better option. The earlier discussed controlled laboratory experiment of Malmendier and Schmidt (2012) involving business gift with a negative externality was designed to rule out any repeated interaction, informational content in the gift or monetary incentives. The results showed that the gift as if still induced reciprocity, even if it was quite small. The experiment gives results that are actually inconsistent with most theories of outcome based social preferences such as altruism and inequity aversion, because if one producer is favored another one is hurt and in this experiment there even is the negative externality that the client is hurt as well (Malmendier and Schmidt, 2012). Also, in their design they rule out intention based reciprocity as a cause, because the ‘’intentions’’ send in their experiment would actually be negative and thus should not be reciprocated. The
explanation that they give for their results is the fact that giving a gift makes the recipient feel obligated to return something, which creates a social bond. This result is then linked to anthropological and sociological literature based on the norm that gifts need to be
reciprocated. Malmendier and Schmidt (2012) suggest an extension to models based on social preferences by incorporating the effect of a reference point. The reference point is then equal to the expected behavior, in which gift giving strengthens the bond between two individuals and increases the reciprocal behavior of the recipient. The more the actions of the gift giver exceed expectations, the higher the reciprocal behavior will be.
Despite the fact that the results of Malmendier and Schmidt (2012) show that in their design intention based reciprocity cannot cause the reciprocal behavior after the receiving of a gift
12 there are multiple studies which show the importance of intentions (e.g. Falk, Fehr,
Fischbacher, 2008; Greenberg & Frisch, 1972; Lowe & Goldstein, 1979). Perhaps if the design of Malmendier and Schmidt (2012) did give the opportunity to send out positive intensions, the results would show that this increases reciprocity. Since a gift is generally given with positive intentions, this could change the results. Even if there remain studies that do not show that intentions matter, they do not take different cultures into account. This could be an important omission. Between cultures, intentions can play a major role and give
different responses.
2.6 Size of the gift
The size of the gift can also lead to different responses. First of all, it should be realized that small gifts can still be of great influence. There are several studies that confirm that gifts with practically no value, such as pens, note pads, and coffee, can still significantly affect behavior (Katz, Caplan & Merz, 2003). This means that even small gifts can have large effects. If something is of little value it can still be highly appreciated or rejected. For
example, if something is perceived as a bribe people become too ashamed to reciprocate (Ong, 2011).
There are some mixed results on if large or small gifts have a bigger effect. The experiment of Malmendier and Schmidt (2012) finds that small gifts have stonger effects than large gifts, since here small gifts favor the gifts giver in 68% of the time and large gifts in 50% of the time. This is in expectation with their theory of social preferences with an exogenous
reference group. The expectations will be higher when the gift benefits them more and when a gift is expected the reciprocal reward is smaller. However, this is the case here because
someone then expects the gift. Unexpected gifts can show different results and this is not taken into account.
A larger size of the gift could thus still be of great importance for the effect that it has. Falk (2007) showed that a small gift increases donations to a charity by 17% while a large gift increased them by 75%. This confirms that even though both gifts have a reciprocal effect, the size of the gift makes a difference. In general it is important to take into account that both outcomes (the size of the gift) and intentions (what someone tries to achieve by giving the gift) need to be taken into account to be able to say something about the differences in the effect between a small and a large gift (Falk & Fischbacher, 2006). If intentions are perceived
13 to be good or bad a larger gift will thus give stronger results by strengthening the outcome effect.
2.7 Cross-cultural differences
It needs to be realized that intentions can be perceived differently when looking at different cultures. Arunthanes, Tansuhai and Lemak (1994) recognize the lack of cross cultural comparisons within business gift-giving. If cross-cultural business gift giving is not done in the right way, this could deteriorate potential new business relations or stop fruitful negotiations (Arunthanes, Tansuhai & Lemak, 1994). However, experiments testing the effect and causes of gift giving are mostly conducted in western countries. Besides this there are to my knowledge no empirical studies which compare cross-cultural differences in gift giving. This gives a gap in the literature and an important possibility for future research.
This study attempts to fill this gap by looking at the cross-cultural differences in the perceived gains and losses after receiving a business gift. The study is executed in both Poland and the Netherlands to be able to see if there are cultural differences that may lead to a gift being experienced differently. Poland is used because it is one of the healthiest and fastest growing post-communist countries in the European Union. It is a high income economy, because of its booming market economy (Rountree, O’Kreefe & Chadraba, 2014). This makes it an
attractive business partner for many Western countries including The Netherlands. This causes the Netherlands to be one of the main investors of Poland and a high trading volume exists between the two countries. However, its post-communist past gives Poland a totally different set of cultural values than the Netherlands (Kolman et al., 2002). Marktadviser Rob Sterkman explains that communist past makes Poland very bureaucratic and leads to a different mentality between the Polish and Dutch population. The Polish people do not necessary think in win-win situations as the Dutch do. Their mentality is more ‘’if you win, than I must be losing’’. He explains that this leads to business gifts not being perceived as ‘’nice, I get more for the same amount of money’’ but as ‘’I get more, so I must have paid too much’’ (‘’Onderneem Polen’’, 2012). This means that business gifts that are given as
appreciation of past negotiations can be experienced in a negative way instead of a positive one, as one might think beforehand.
There are several studies that confirm the cultural differences between post-communist Poland and the West, for example Kolman and others (2002) compare it on Hofstede’s five
14 dimensions with the Netherlands. Hofstede’s five dimensions of national cultures are the attempt to capture the impact that cultural values can have in the workplace. Hofstede (1980) used survey data from 50 countries to investigate the cultural differences on several
dimensions. These dimensions are: power distance, individualism-collectivism, uncertainty avoidance, masculinity-femininity and long versus short-term orientation. From which the last dimension was not part of the original study, but added later on (Hofstede & Bond, 1988). The study of Kolman and others in 2002 shows that Poland and the Netherlands score relatively different on the five dimensions, indicating that large cultural differences are still shown to exist (Kolman et al., 2002) This is important to know, since this can lead to different responses to a business gift. The dimension that could be particularly important for this study about business gifts is individualism-collectivism, because it has been show that this
dimension is related to negotiation behavior. Individualism-collectivism measures to what extent a certain culture relies on and has loyalty to the individual or to the group. More collectivistic cultures believe that there is a great distance between members of the ‘’group’’ and outsiders. While individualistic cultures do not perceive this distance to exists (Hofstede, 1980). Poland is much more collectivistic than The Netherlands (Kolman et al., 2002). Management literature confirms this finding and shows that Poles more often have to
tendency to work closely together in ‘’beating the enemy’’ (Yanouzas & Boukis, 1993). This behavior most certainly has something to do with the former communist values of shared responsibility, but these collectivists values are likely to be rooted more strongly within the Polish culture. This can also be seen in the fact that when doing business in Poland, it is important to build a relationship with your trading partner first. Eating and drinking together are an important aspect of this. Poles believe that business can only proceed smoothly after you have built a good relationship with someone. This results from a certain ‘’culture of distrust’’ that is present in all levels of Polish social life (Sztomka, 1996). In the example of Rob Sterkman mentioned above (‘’Onderneem Polen’’, 2012) this distrust has taken a form against the goodwill of the trading partner after receiving a gift.
It is important to understand more about these differences that affect business negotiations between post-communist countries and the West. In a clear-cut example of comparing Poland and The Netherlands in their responses to a hypothetical situation that involves the receiving of a gift these differences become less abstract than the broad cultural differences that are shown to exist (Kolman et al., 2002). This can lead to more insights that can readily be used in real-life negotiations such that business men trading with Poland know what to pay
15 attention to. Furthermore it can help to find implementations for further research in this area, such that potential results can be extended to other countries.
16 3. Methodology
3.1 Sample
Concerning the sample it would be best to have data from real Polish and Dutch business men, however these are generally not eager to cooperate and difficult to reach. Furthermore, the sample would than differ on many different aspects (i.e. age, size of business). This is why a survey with students is used, similar to the method used by Kolman and others (2002). In the experiment of Kolman and others (2002) five different countries (including Poland and the Netherlands) were compared by surveying university students who had the nationality of the country of interest. The strategy of matched samples (Hofstede, 1991) was used, which means that the participants used were similar in as many aspects as possible (i.e. university students), instead of trying to have a sample that represents the whole population of a country. Using the strategy of matched samples leads to many possible sources of variation to be cancelled out and results in a sample that is comparable in many respects besides nationality.
The final sample in this experiment consisted of 233 participants, but 45 subjects had to be deleted before starting the analyses. This was done because of several reasons: 14 were deleted because they failed to fill in their demographics or stopped at any point earlier in the experiment; 23 since they stated that studying was not their main occupation; 8 because of nationality reasons such as not growing up in the country of their nationality, both parents not being of the same nationality or having a completely different nationality than Dutch or Polish. This resulted in a final sample of 188 participants, from which 97 are Dutch and 91 are Polish. The sample is dominated by females by having approximately 30% male and 70% of female participants. However, this is the case for both nationalities. If there thus is a difference in gender, this will be the case in both countries. Also, gender is controlled for when testing the significance levels. The age of the participants ranged from 17 to 30 years old, but the largest part of the sample (81.38%) was 19 to 23 years old.
Total Poland Netherlands
N 188 91 97
Male 30.32% 30.77% 29.90%
Female 69.68% 69.23% 70.10%
17 To be able to make sure there are no differences in responses caused by the field of study between the two countries, the field of study was also asked to the participants. Even though not all students filled this in, the field of study in both countries was largely varied for the students who did. This rules out that the participants answered differently between countries because, for example, in one country a majority of economics students are used and in the other psychology students. Also, the fact that the survey was spread among students of different institutions and in varying places (e.g. canteen, library, study center) helps prevent this.
3.2 Experimental design
3.2.1 Vignette method
For this study the Vignette method is used with students in both Poland and the Netherlands. The method of study with vignettes means that ‘’short descriptions of a certain person or social situation that contain precise references to what are thought to be the most important factors in the decision-making or judgment-making processes of respondents’’ (Alexander & Becker, 1978) are provided to all students.
3.2.2 Components of the experiment
The experiment itself will take the following form. First a business negotiation involving the receiving of a business gift will be sketched without information about the deal being a zero-sum game or a win-win situation. Several questions will then be asked to find out how they indicate their position in the deal after they received the gift. These questions will involve: how satisfied they feel about the deal and the receiving of the present; their perceived likelihood of doing business with this person again and how acceptable they believe the value of the gift is. After they answer these questions more information will be provided about the division of the gains from the trade, either as a zero-sum or a win-win situation. After this, the same questions to elicit how they perceive their position in the deal are asked again. In the end questions related to their demographics are asked. Besides the general age and gender questions, there were several questions related to their nationality. It was
important to know more than just the nationality itself, such as if they grew up in the country of their nationality and if both of their parents are of the same nationality. For example, there was one subject in the sample who had the Dutch nationality, but a Russian parent. Since this
18 parent with a communist past could have influenced his way of thinking more towards the Polish mindset, he cannot be used in the sample as a Dutch participant. The last questions asked the participants about their main occupation, this was to rule out people who were not a full-time student. This was necessary to be able to use the strategy of matched samples to its best.
Business negotiation without extra information Questions round 1
Extra information about zero-sum or win-win Questions round 2
Demographics
Table 2: Components of survey
For both the zero-sum and win-win treatment there is a version with a small and a large gift. This gives a total of four different versions of the survey. The small gift is represented by a bottle of wine worth €10 and the large gift as an Apple iPad worth €500. These are considered small and large in comparison to the total value of the deal which is €10,000
To increase data collection, both an online and printed version of the survey was spread among students. The complete surveys can be found in the Appendix.
Zero-sum treatment Win-win treatment
Small gift (1) Small gift (3)
Large gift (2) Large gift (4)
Table 3: The different versions of the survey
3.2.3 Type of answer possibilities
For all answers possibilities, leaving aside the questions related to demographics, seven-level Likert scales are used. The answer possibilities range from the most negative option at 1 to the most positive option at 7. This gives the participants the possibility to answer neutrally in the middle if they desire this.
3.2.4 Data analysis
A regression analysis is performed to test if there are differences in the response to the additional information between Polish and Dutch participants. The difference in the answers before and after additional information (dependent variable) is regressed on a nationality
19 dummy (independent variable). The regression coefficient then gives the
difference-in-differences. This coefficient thus is the difference between nationalities in the difference before and after additional information. To analyze if there are any differences in the
participants response between gift-sizes the regression is performed with a gift-size dummy as independent variable. This coefficient then gives the difference in responses between small and large gifts in the difference before and after additional information.
In these regression analyses gender is controlled for, by adding a gender dummy to the regression. This is done, since the sample is female-dominated by having approximately 70% female respondents. The adding of gender in the regression leads to the regression coefficients of nationality and gift-size to represent the mean change in the answers between nationality and gift-size to the response variables: difference (before and after additional information) of satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift), while holding gender constant (ceteris paribus).
The p-value accompanied by the regressions then gives the significance level of these difference-in-differences, at either 1% (p<0.01), 5% (p<0.05), or 10% (p<0.10). A low p-value indicates that the null hypothesis of the coefficient being equal to zero can be rejected. For example, a p-value of less than 5% indicates that the chances of finding a value equal to the coefficient, while it was actually 0 is less than 5% (Stock & Watson, 2012).
3.3 Hypotheses
Based on the previously discussed literature several hypotheses can be formed. Intentions matter for reciprocal behavior (Falk & Fischbacher, 2006) and can be perceived differently when looking at different cultures (Arunthanes, Tansuhai & Lemak, 1994). It has been shown that Poland and the Netherlands have a different set of cultural values (Kolman et al., 2002). One of the important differences is that in Poland they tend to be more
collectivistic than in the Netherlands (Kolman et al., 2002) and feel as if they need to work together in ‘’beating the enemy’’ (Yanouzas & Boukis, 1993). There is a general culture of distrust rooted within Polish culture. Moreover, the explanations of Rob Sterkman
(‘’Onderneem Polen’’, 2012) can lead to the prediction that the difference between Polish and Dutch people to receiving a gift can be found in how they perceive the division of the gains from trade. Dutch people think more in win-win situations, while Poles see the trade as a
20 zero-sum game, meaning that if one person wins the other loses. This leads to the first two hypotheses.
Zero-sum treatment
Hypothesis 1: Polish will answer the questions more similar in both rounds (already perceived it as zero-sum) compared to the Dutch who will answer them more negatively in the second round (perceived it as win-win before they got extra information).
Win-win treatment
Hypothesis 2: Dutch will answer the questions more similar in both rounds (already perceived it as win-win) compared to the Poles who will answer them more positively in the second round (perceived it as zero-sum before they got extra information).
Concerning the size of the gift we can form the final hypothesis. Even though small gifts can still be of great influence (Katz, Caplan & Merz, 2003) the size of the gift can strengthen the reciprocal effect (Falk, 2007). Both the intentions about what someone tries to achieve by giving the gift and the outcome in the form of the value of the gift matter for the size of the reciprocal effect of the receiver (Falk & Fischbacher, 2006). So if the intentions are perceived in a certain way (e.g. good or bad) a larger gift will give stronger responses by strengthening the outcome effect. This gives the following hypothesis.
Small versus large gifts
21 4. Results
4.1 Summary statistics: means of outcome variables
In order to answer hypotheses 1 and 2 the levels of: how satisfied they feel about the deal and the receiving of the present; their perceived likelihood of doing business with this person again and how acceptable they believe the value of the gift is, are asked before and after the additional information. In table 4 and graph 1 the average values of the answers given before any additional information was provided are shown, as are the values after the additional information for both the zero-sum and win-win treatment. These values are shown separately for the Dutch and Polish participants.
To be able to answer hypothesis 3 the answers to the questions must be separated for the participants who received version with a small gift or a large gift. The average values per nationality and gift-size can be seen in table 5 and graph 2. While, the average values with pooled nationalities (so only distinguishing in gift-size) can be found in table 6 and graph 3.
22
Before additional information After zero-sum After win-win
Polish Dutch Polish Dutch Polish Dutch
N 91 97 44 49 47 48
Satisfaction with deal Mean 5.341 5.577 4.432 4.225 4.574 5.604
(SD) 1.352 1.189 1.453 1.279 1.426 1.086
Business again Mean 5.395 5.567 4.659 4.735 4.617 5.458
(SD) 1.182 0.978 1.627 1.319 1.261 1.031
Satisfaction with gift Mean 4.923 5.433 4.409 4.449 4.234 5.458
(SD) 1.432 1.249 1.647 1.292 1.386 0.988
Value of gift Mean 4.626 4.887 4.114 4.408 4.383* 4.896
(SD) 1.717 1.322 1.646 1.413 1.662 1.259
The means shown in this table are the average values of the answers to the questions related to the participants’: satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift. The answer possibilities ranged from 1 (most negative) to 7 (most positive). The table separates the means for Polish and Dutch participants and shows the means before additional information, after zero-sum additional information, and after win-win additional information.
23 The means shown in this graph are the average values of the answers to the questions related to the participants’: satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift. The answer possibilities ranged from 1 (most negative) to 7 (most positive). The graph separates the means for Polish and Dutch participants and shows the means before additional information, after zero-sum additional information, and after win-win additional information.
Graph 1: Means of the outcome variables per nationality before and after any additional information is given. 0 1 2 3 4 5 6 7
Before info Polish Before info Dutch Zero-sum Polish Zero-sum Dutch Win-win Polish Win-win Dutch
Satisfaction with deal Business again Satisfaction with gift Value of gift
24
Before additional information After zero-sum After win-win
Small gift Large gift Small gift Large gift Small gift Large gift
Polish Dutch Polish Dutch Polish Dutch Polish Dutch Polish Dutch Polish Dutch
N 43 48 48 49 21 23 23 26 22 25 25 23
Satisfaction with deal Mean 5.721 5.604 5.000 5.551 4.381 4.217 4.478 4.231 4.591 5.720 4.560 5.478
(SD) 0.908 1.180 1.584 1.209 1.203 1.278 1.675 1.306 1.333 0.980 1.530 1.201
Business again Mean 5.372 5.750 5.417 5.388 4.667 4.435 4.652 5.000 4.455 5.720 4.760 5.174
(SD) 1.113 0.812 1.252 1.096 1.426 1.502 1.824 1.095 1.101 0.614 1.393 1.302
Satisfaction with gift Mean 5.140 5.583 4.729 5.286 4.429 4.391 4.391 4.500 4.273 5.600 4.200 5.304
(SD) 1.187 0.986 1.608 1.458 1.363 1.305 1.901 1.304 1.162 0.866 1.581 1.105
Value of gift Mean 4.884 4.917 4.396 4.857 3.905 4.087 4.304 4.692 4.773 5.120 4.040 4.652
(SD) 1.401 1.200 1.943 1.443 1.446 1.240 1.812 1.517 1.478 1.166 1.767 1.335
The means shown in this table are the average values of the answers to the questions related to the participants’: satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift. The answer possibilities ranged from 1 (most negative) to 7 (most positive). The table separates the means for both nationality and gift-size and shows the means before additional information, after zero-sum additional information, and after win-win additional information.
25 The means shown in this graph are the average values of the answers to the questions related to the participants’: satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift. The answer possibilities ranged from 1 (most negative) to 7 (most positive). The graph separates the means for both nationality and gift-size and shows the means before additional information, after zero-sum additional information, and after win-win additional information.
Graph 2: Means of the outcome variables per nationality before and after any additional information is given separated for large and small gifts 0 1 2 3 4 5 6 7
Satisfaction with deal Business again Satisfaction with gift Value of gift
26
Before additional information After zero-sum After win-win
Small gift Large gift Small gift Large gift Small gift Large gift
N 91 97 44 49 47 48
Satisfaction with deal Mean 5.659 5.278 4.295 4.347 5.191 5.000
(SD) 1.056 1.427 1.231 1.480 1.279 1.444
Business again Mean 5.571 5.402 4.545 4.837 5.128 4.958
(SD) 0.979 1.170 1.454 1.477 1.076 1.352
Satisfaction with gift Mean 5.373 5.010 4.409 4.449 4.979 4.729
(SD) 1.102 1.551 1.317 1.595 1.207 1.469
Value of gift Mean 4.901 4.629 4.000 4.510 4.957 4.333
(SD) 1.291 1.716 1.329 1.660 1.318 1.589
The means shown in this table are the average values of the answers to the questions related to the participants’: satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift. The answer possibilities ranged from 1 (most negative) to 7 (most positive). The table separates the means for gift-size, but combines them for both nationalities. It shows the means before additional information, after zero-sum additional information, and after win-win additional information.
27 The means shown in this graph are the average values of the answers to the questions related to the participants’: satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift. The answer possibilities ranged from 1 (most negative) to 7 (most positive). The graph separates the means for gift-size, but combines them for both nationalities. It shows the means before additional information, after zero-sum additional information, and after win-win additional information.
Graph 3: Means of the outcome variables per gift-size for all participants before and after any additional information is given 0 1 2 3 4 5 6 7
Before info small gift Before info large gift Zero-sum small gift Zero-sum large gift Win-win small gift Win-win large gift
Satisfaction with deal Business again Satisfaction with gift Value of gift
28 4.2 Difference-in-differences
To be able to say anything about the hypotheses it needs to be compared if the
difference between the answers before and after additional information is significantly different between nationalities or gift-size.
4.2.1 Difference-in-differences between nationalities
The differences (value after additional information minus value before additional information) per treatment and nationality are shown in table 7 and graph 4. It is then analyzed if these differences are significantly different per nationality. If a question is answered more negatively (positively) after the additional information, the mean will be more negative
(positive). As can be seen only the satisfaction with the deal and gift are significantly different between nationalities in the zero-sum treatment. While, in the win-win treatment all variables are significantly different between nationalities. In the zero-sum treatment the means of both the Polish and Dutch are negative, which means that they answered more negatively after the additional information. In the win-win treatment the Poles remain to have answered more negatively after the additional information, while the Dutch switch so a positive mean, indicating that they answered more positively after the additional information.
29 Table 7: Difference-in-differences before and after additional information per nationality
Total
Zero-sum Win-win
Difference in Polish Dutch Difference Polish Dutch Difference
N 44 49 47 48
Satisfaction with deal Mean -0.727 -1.429 0.701** -0.936 0.104 -1.040***
(SD) 1.648 1.472 1.358 1.341
Business again Mean -0.818 -1.020 0.202 -0.702 0.083 -0.785***
(SD) 1.280 1.200 1.012 1.069
Satisfaction with gift Mean -0.636 -1.122 0.486* -0.574 0.167 -0.741**
(SD) 1.296 1.269 1.058 1.730
Value of gift Mean -0.545 -0.673 0.128 -0.213 0.208 -0.421*
(SD) 1.190 1.297 0.999 1.129
*, **, *** indicate p<0.1, <0.05, <0.01 in the linear regression controlling for gender
The means shown in this table are the difference (value after additional information minus value before additional information) per variable (satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift). The table separates the means for Polish and Dutch participants and shows the difference for the zero-sum and win-win additional information. The ‘’Difference’’ column then shows the difference-in-differences between nationalities and the significance of this difference-in-differences.
30 The means shown in this graph are the difference (value after additional information minus value before additional information) per variable (satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift). The graph separates the means for Polish and Dutch participants and shows the difference for the zero-sum and win-win additional information.
Graph 4: Difference-in-differences before and after additional information per nationality -2 -1,5 -1 -0,5 0 0,5 1
Zero-sum Polish Zero-sum Dutch Win-win Polish Win-win Dutch
Satisfaction with deal Business again Satisfaction with gift Value of gift
31 4.2.2 Difference-in-differences between nationalities and gift-size
In table 8 the separation between the small and large gift per nationality is made again in. In the zero-sum treatment none of the variables are significantly different per nationality for the small gift version. For the large gift version satisfaction with the deal and satisfaction with the gift are significantly different between nationalities again. In the win-win treatment the variables are significantly different per nationality for both the small and large gift version, with the exception of the satisfaction with the gift and the acceptability of the value of the gift in the large gift version. In graph 5 it is clearly seen that the Poles always answer more
negatively after the additional information, while the Dutch tend to answer more positively in the win-win treatment.
32 Table 8: Difference-in-differences before and after additional information per nationality separated for large and small gifts
Zero-sum Win-win
Small gift Large gift Small gift Large gift
Difference in Polish Dutch Difference Polish Dutch Difference Polish Dutch Difference Polish Dutch Difference
N 21 23 23 26 22 25 25 23
Satisfaction with deal Mean -1.190 -1.304 0.114 -0.304 -1.538 1.234*** -1.273 0.040 -1.313*** -0.640 0.174 -0.814*
(SD) 1.692 1.295 1.521 1.630 1.386 1.136 1.287 1.557
Business again Mean -0.810 -1.261 0.451 -0.826 -0.808 -0.018 -0.818 -0.080 -0.738*** -0.600 0.261 -0.861**
(SD) 1.167 1.176 1.403 1.201 1.053 0.640 1.000 1.389
Satisfaction with gift Mean -0.857 -1.087 0.230 -0.435 -1.154 0.719** -0.727 -0.080 -0.647** -0.440 0.435 -0.875
(SD) 1.389 1.535 1.199 1.008 0.985 0.997 1.121 2.273
Value of gift Mean -0.714 -0.870 0.155 -0.391 -0.500 0.109 -0.364 0.240 -0.604* -0.080 0.174 -0.254
(SD) 1.454 1.217 0.891 1.364 1.049 1.200 0.954 1.072
*, **, *** indicate p<0.1, <0.05, <0.01 in the linear regression controlling for gender
The means shown in this table are the difference (value after additional information minus value before additional information) per variable (satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift). The table separates the means for Polish and Dutch participants and gift-size. The table shows the difference for the zero-sum and win-win additional information. The ‘’Difference’’ column then shows the difference-in-differences between nationalities and the significance of this difference-in-differences.
33 The means shown in this graph are the difference (value after additional information minus value before additional information) per variable (satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift). The graph separates the means for Polish and Dutch participants and gift-size. The table shows the difference for the zero-sum and win-win additional information.
Graph 5: Difference-in-differences before and after additional information per nationality separated for large and small gifts -2 -1,5 -1 -0,5 0 0,5 1
Satisfaction with deal Business again Satisfaction with gift Value of gift
34 4.2.3 Difference-in-differences between gift-size
In table 9 the difference in the answers before and after additional information can be seen per treatment and gift-size, but with the nationalities pooled together. Just as in the analyses with separate nationalities the answers are always more negative after receiving additional information in the zero-sum treatment. However, without the distinction between nationalities the win-win treatment now also shows only negative numbers, except for the value of the gift in the large gift version. This is also shown in graph 6. However, when comparing the difference-in-differences between small and large gifts it can be seen that the difference is no significantly different between small land large gifts for any of the variables in both treatments.
35 Table 9: Difference-in-differences before and after additional information per gift-size
Total
Zero-sum Win-win
Difference in Small gift Large gift Difference Small gift Large gift Difference
N 44 49 47 48
Satisfaction with deal Mean -1.25 -0.959 -0.291 -0.574 -0.250 -0.324
(SD) 1.48 1.6827 1.410 1.466
Business again Mean -1.045 -0.816 -0.229 -0.426 -0.188 -0.238
(SD) 1.18 1.286 0.927 1.265
Satisfaction with gift Mean -0.977 -0.816 -0.161 -0.383 -0.021 -0.362
(SD) 1.455 1.149 1.033 1.804
Value of gift Mean -0.795 -0.449 -0.346 -0.043 0.417 -0.460
(SD) 1.322 1.156 1.160 1.010
*. **. *** indicate p<0.1, <0.05, <0.01 in the linear regression controlling for gender
The means shown in this table are the difference (value after additional information minus value before additional information) per variable (satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift). The table separates the means for gift-size, but combines them for both nationalities. The table shows the difference for the zero-sum and win-win additional information. The ‘’Difference’’ column then shows the difference-in-differences between gift-size and the significance of this difference-in-differences.
36 The means shown in this graph are the difference (value after additional information minus value before additional information) per variable (satisfaction with the deal, willingness to do business again, satisfaction with the gift, and how they feel about the value of the gift). The graph separates the means for gift-size, but combines them for both nationalities. The graph shows the difference for the zero-sum and win-win additional information.
Graph 6: Difference-in-differences before and after additional information per gift-size -2 -1,5 -1 -0,5 0 0,5
Zero-sum small gift Zero-sum large gift Win-win small gift Win-win large gift
Satisfacation with deal Business again Satisfaction with gift Value of gift
37 4.3 Possible limitations
4.3.1 Validity of vignette method
There are some positive and negative aspects about using the Vignette method. Concerning the internal validity the method provides standardized stimuli to all the
participants, however the external validity is weaker since the incentives are low. Moreover, the vignette can only summarize the thoughts of the participants, since they have limited answer possibilities. This means that the underlying thoughts to the answers cannot be elicited, but only hypothesized.
4.3.2 Nationality
It could be a problem if someone states in the survey that he is from a certain
nationality, but did not grow up in the country or was not raised by parents of that nationality. Potential biases related to the persons’ nationality were taken into account by deleting certain subjects. Their nationality was controlled for by checking if both of the participants’ parents were from the same nationality and if they grew up in the country of their nationality. Still there could be other influences on someone’s nationality that were no controlled for.
4.3.3 Strategy of matched samples
Another limitation is that subjects can answer differently between countries, because of other reasons than their nationality (e.g. occupation, field of study). To prevent this as much as possible the strategy of matched samples was used to make sure that the participants were similar on many aspects. This is why students who did not report studying as their main occupation were subsequently deleted from the analyses. Also, the field of study was recorded for a subsample and was largely varied. Moreover, the survey was spread at many different locations, both online and by hand. This was to make sure a varied sample of students was used in both countries, to rule out any biases related to the field of study.
4.3.4 Sample not representative
In this experiment students are used, instead of real businessmen. This could mean that the sample is not representative for the people who actually receive business gifts.
38 experiment. Also, using students means that the strategy of matched samples could be used. This shows that even though using students could be a limitation, it was still a reasonable choice for this experiment.
4.3.5 Language problems
While preparing for data collection it was brought to my attention that Polish students might have more trouble understanding an English survey than Dutch students. Since it would be a problem if participants filled in the survey, but did not completely understand the texts or questions, several measures were taken. First, words that could beforehand be experienced as difficult were given one or multiple synonyms behind them in brackets. As an example it was stated in the survey that the negotiation was tough (hard / not easy). The full survey in the Appendix shows all other words for which extra explanations were given. Second, when surveys were handed out manually it was told to the participants that if they did not
understand anything they could let me know. If this was the case either an extra explanation was given or their survey was not used. For the online survey it was mentioned when
spreading that the survey was in English and that basic English knowledge was required. An extra control was the fact that the online survey started with an introduction with quite a lot of text and this can be seen as the most difficult part of the survey English wise. It is assumed that if participants understand this introduction they will understand the rest of the survey and that if they did not understand the introduction they will choose not to continue and their answers will not be recorded.
39 5. Discussion and conclusion
5.1 Main findings
5.1.1 Difference-in-differences per treatment
It was expected in hypothesis 1 and 2 that Polish participants will have perceived the deal as zero-sum to start with and Dutch participants will see it as win-win before they received additional information. In the zero-sum treatment it was then hypothesized that the Polish will answer the questions more similar in both rounds (already perceived is as zero-sum) compared to the Dutch who will answer the more negatively in the second round (perceived it as win-win before they got extra information). The opposite reasoning holds for the win-win treatment. Here it was hypothesized that the Dutch will answer the questions more similar in both rounds (already perceived it as win-win) compared to the Poles who will answer them more positively in the second round (perceived it as zero-sum before they got extra information).
When looking at the difference-in-differences from table 7 and 8 it needs to be realized that the answers are presented on a seven-level Likert scale ranging from the most negative option at 1 to the most positive option at point 7. If answers are thus changed to be more negative (positive) after the additional information, the mean of the difference (answers 2nd round – answers 1st round) will become negative (positive).
This means that in the zero-sum treatment it is predicted by hypothesis 1 that the difference in answers for the Dutch is larger (but as a negative number) than the difference in answers for the Polish (also having a negative number): (Answer 2nd round Dutch - Answer 1st round Dutch) |>| (Answer 2nd round Polish – Answer 1st Round Polish) or Difference Dutch > Difference Polish. This leads to the expectation of positive difference means, since these are calculated as Difference Polish – Difference Dutch. Table 7 confirms this expectation by having all positive numbers as difference means for the zero-sum treatment. However, this is only significantly different between nationalities for 2 out of the 4 variables, namely
satisfaction with the deal and satisfaction with the gift.
For the win-win treatment it is expected that the difference in answers for the Polish is larger than the difference in answers for the Dutch: (Answer 2nd round Polish – Answer 1st Round Polish) > (Answer 2nd round Dutch - Answer 1st round Dutch) or Difference Polish >
40 Difference Dutch. This leads to the expectation of negative difference means, since these are calculated as Difference Polish – Difference Dutch. Table 7 rejects the expectations, even though we do find all significant negative difference means. Nevertheless, when looking more closely at the numbers for the Polish and Dutch separately (table 7 and graph 4) it can be seen that this is the case because only the Dutch have positive means (i.e. answered more
positively after additional information) while the Poles have answered more negatively after the additional information. This is against the predictions of hypothesis 2. Not only do the Dutch answer more positively than the Poles, the Poles even answer the questions more negatively after learning that the negotiation is a win-win situation.
To get information on hypothesis 3, which predicts that the response to large gifts will be stronger compared to small gifts, the separation of the results from table 8 and graph 5 is made for the small and large gift version of the survey. To be able to formally test hypothesis 3 and find out whether large gifts give stronger responses than small gifts the difference-in-differences analyses of gift-size in table 9 was performed with pooled nationalities.
In table 8 it can already be seen that in the zero-sum treatment results do not appear to be stronger for the large gift version than for the small gift version. In some cases the difference increases while in others it decreases compared to the analyses when there was no separation with the large or small gift. It can be seen that none of the differences between Polish and Dutch participants are significant for the small gift version, while 2 out of 4 are significant for the large gift version. This is equal to the 2 out of 4 significant results when there was no separation.
Table 9 then shows that with pooled nationalities there are actually no significant differences between large and small gifts in the zero-sum treatment.
In the win-win treatment the separation in table 8 between small and large gifts does not change anything about the fact that the results remain opposite as expected with Dutch becoming more positive after the additional information and Poles turning negative. Nonetheless, the separation of gift size does reveal something interesting. In the large gift version, both the Polish and Dutch participants become more positive after the additional information in all variables except the acceptability of the value of the gift in which only the Polish become more positive, compared to the situation where there was no gift separation. In the small gift version both become less positive than without the gift separation in all
