{\rtf1\ansi\ansicpg1252\deff0 {\fonttbl {\f0\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f1\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f2\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f3\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f4\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f5\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f6\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f7\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f8\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f9\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f10\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f11\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f12\fnil\fcharset0\fprq0\fttruetype Symbol;} {\f13\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f14\fnil\fcharset0\fprq0\fttruetype Symbol;} {\f15\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f16\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f17\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f18\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f19\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f20\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f21\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f22\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f23\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f24\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f25\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f26\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f27\fnil\fcharset0\fprq0\fttruetype Tms Rmn;} {\f28\fnil\fcharset0\fprq0\fttruetype Tms Rmn;} {\f29\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f30\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f31\fnil\fcharset0\fprq0\fttruetype Times New Roman;} {\f32\fnil\fcharset0\fprq0\fttruetype Times New Roman;}} {\colortbl \red0\green0\blue0;} \kerning0\cf0\viewkind1\paperw12240\paperh15840\margl1440\margr1440\widowctl \sectd\sbknone\colsx360 \pard\qc\ri180{\f0\fs28\b Theories of Commitment, Altruism and Reciprocity: Evidence from Linear Public Goods Games} \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180{\f1 Rachel T.A. Croson} \par\pard\qc\ri180{\f1 CrosonR@opim.wharton.upenn.edu} \par\pard\qc\ri180{\f2\fs20 April, 1998} \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180 \par\pard\qc\ri180{\f3\b Abstract} \par\pard\qc\ri180 \par\pard\ri180{\f4 Theories of commitment, altruism and reciprocity have all been invoked to explain and describe observed behavior in public goods and social dilemma situations. In particular, commitment theories have been used to explain behaviors like water conservation and voting. Theories of altruism are applied in explanation of contributions to charities and intergenerational transfers and bequests. And theories of reciprocity have been invoked to explain gift exchange and labor market decisions. This paper describes a set of experiments which distinguish between these competing theories by testing their comparative statics predictions in a linear public goods setting. Results provide strong support for reciprocity theories over either theories of commitment or of altruism.} \par\pard\ri180 \par\pard\ri180{\f4 Keywords: Reciprocity, Altruism, Experimental economics, Public goods, Charitable contributions} \par\pard\ri180 \par\pard\ri180{\f4 JEL Classification Codes: C9, D64, H41, C72} \par\pard\ri180 \par\pard\ri180 \par\pard\ri180 \par\pard\ri180 \par\pard\ri180{\f4 ________________________} \par\pard\ri180{\f5\fs20 *The author thanks Jon Baron, Yan Chan, Robyn Dawes, Jerry Green, Elizabeth Hoffman, Mark Isaac, Eric Maskin, Sara Solnick, Lise Vesterlund, George Wu, participants at the Economic Science Association Fall Meetings as well as seminar participants at the University of Pennsylvania and the University of Iowa for helpful comments. All omissions or mistakes are the responsibility of the author. Funding of experiments from the Economic Science Lab at the University of Arizona is gratefully acknowledged.} \par\pard\ri180\sl480\slmult1{\f6\fs20\page}{\f7\b 1. Introduction} \par\pard\ri180\sl480\slmult1{\f8\tab US individuals made over 100 billion dollars of philanthropic contributions in 1995 (}{\f9\i Giving USA,}{\f8 1996). This behavior is inconsistent with traditional utility theory in which individuals care only for their own consumption. A number of alternative theories have been invoked to explain such philanthropic behavior in this and other settings. This paper describes a set of experiments which distinguish between three competing theories: commitment, altruism and reciprocity, by testing their comparative statics predictions in a linear public goods setting.} \par\pard\ri180\sl480\slmult1\tx720{\f10\tab In commitment theories, individuals choose the actions they would most prefer everyone would choose (Laffont, 1975; Harsanyi, 1980). Thus they choose the action which maximizes their private payoff assuming that everyone else chooses the same action they do. Commitment theories are consistent with observed philanthropic behavior, voluntary cooperation in social dilemmas like water conservation (Laffont, 1975), tax evasion (Baldry, 1987), and voting (Struthers and Young, 1989) as well as voluntary contributions to public goods.} \par\pard\ri180\sl480\slmult1{\f8\tab In altruism theories, the consumption of others appears positively as an argument in an individual\uc0\u8217 s utility function (Becker, 1974; Andreoni, 1989, 1990). Models of altruism are also consistent with observed philanthropic behavior and have been used to explain intergenerational bequests (Becker, 1974), social security and other welfare systems (Coate, 1995), and helping behavior in the workplace (Rotemberg, 1994) as well as voluntary contributions to public goods.} \par\pard\ri180\sl480\slmult1{\f8\tab In contrast, Sugden (1984) proposes a theory in which the principle of reciprocity acts as a constraint on traditional individual utility maximization. The principle says (roughly) that an individual may not free, cheap or easy ride when others are contributing. Models of reciprocity are also consistent with observed philanthropic behavior (when others are contributing) and have been used to explain individual behavior in tax evasion (Bordignon, 1993), helping in the workplace (Frey, 1993) and labor markets (Akerlof, 1982, 1984; Fehr and Gachter, forthcoming) as well as voluntary contributions to public goods.} \par\pard\ri180\sl480\slmult1{\f7\b\tab}{\f8 This paper presents four separate experiments designed to distinguish between these theories by comparing their comparative statics predictions. The results of the first experiment (presented in section 4) demonstrate a significant and positive relationship between an individual's own contribution and his beliefs of the contributions of others in his group, consistent with theories of reciprocity and inconsistent with traditional self-interested theories or theories of commitment or altruism. The second and third experiments (presented in section 5) test the robustness of the first experiment by comparing an individual\uc0\u8217 s own contributions and the actual contributions of others in his group in different settings. Similar results are generated. The final experiment (presented in section 6) further investigates the specific }{\f9\i type}{\f8 of reciprocity our subjects demonstrate. We find evidence for }{\f9\i middle }{\f8 reciprocity, where players try to match the middle (or average) contribution of the rest of his group, rather than the minimum or maximum.} \par\pard\ri180\sl480\slmult1{\f7\b\tab}{\f8 This paper is organized as follows. Section 2 briefly describes the public goods production function and the voluntary contribution mechanism used in this experiment. Section 3 outlines the three classes of theories and their implications. In section 4 we present the experiment and results designed to distinguish between the competing theories. Section 5 describes two additional experiments designed to test for the robustness of our results. Section 6 describes another experiment which investigates individual behavior in more detail. Finally, section 7 concludes.} \par\pard\ri180\sl480\slmult1 \par\pard\ri180\sl480\slmult1{\f7\b 2. Pure Public Goods and the Voluntary Contribution Mechanism} \par\pard\ri180\sl480\slmult1{\f7\b\tab}{\f8 Pure public goods are goods that are both }{\f9\i nonrival}{\f8 and }{\f9\i nonexcludable.}{\f8 The experiments described in this paper use a linear and pure public good to distinguish between our competing hypotheses. The mechanism used to fund the public good is the voluntary contribution mechanism which most closely parallels philanthropic giving or contributing behavior. This mechanism has been examined extensively in previous literature (see Davis and Holt, 1994, chapter 6 and Ledyard, 1995 for complete reviews).} \par\pard\ri180\sl480\slmult1{\f9\i A. The Mechanism} \par\pard\ri180\sl480\slmult1{\f8\tab The mechanism is structured as follows. Assume each player }{\f9\i i}{\f8 in a group of N identical players has some endowment E}{\f11\fs12 i}{\f8 which can either be contributed to a group account and used to produce units of a public good or can be privately consumed. Call the amount contributed to the group account by }{\f9\i i}{\f8 ,}{\f9\i }{\f8 x}{\f11\fs12 i}{\f8 . The individual\uc0\u8217 s earnings from private consumption is simply the amount consumed (E}{\f11\fs12 i}{\f8 - x}{\f11\fs12 i}{\f8 ). The individual\uc0\u8217 s earnings from contributions to the group account is a multiple of the sum of contributions by all participants in the group P(}{\f12\uc0\u-4013}{\f11\fs12 iN}{\f8 x}{\f11\fs12 i}{\f8 ). } \par\pard\ri180\sl480\slmult1{\f8\tab There is a pure public goods problem whenever
0, and thus }{\f14\uc0\u-3996}{\f16\fs20 x}{\f13\sub i}{\f16\fs20 *}{\f10 /}{\f14\uc0\u-3996}{\f10 x}{\f13\sub j}{\f10 = 0 j\uc0\u8800 iN.} \par\pard\ri180\sl480\slmult1\tx720{\f10\tab So commitment theories have two important implications which we can test in a public goods setting, yielding the}{\f15\i commitment hypothesis}{\f10 . First, they predict strictly positive (but constant) levels of contribution. Second, (the comparative static prediction) they predict a zero correlation between one's contributions and the contributions of others. In particular, under commitment theories each individual chooses the level of contributions which they prefer everyone would choose. As the actual contribution level of others changes, one\uc0\u8217 s own contribution remains stable. Notice, this is the same comparative static prediction as generated by the benchmark theory of self-interest above. Later in our statistical analyses we will look to the absolute level of contributions to distinguish these theories.} \par\pard\ri180\sl480\slmult1\tx720{\f15\i B. Altruism Theories} \par\pard\ri180\sl480\slmult1{\f8\tab A second set of theories of altruism assume that individuals care directly about the consumption or utility of others. These theories then go on to generate behavior which involves (1) positive levels of contributions to public goods, but also (2) contributions which are negatively related to the contributions of others.} \par\pard\ri180\sl480\slmult1{\f8\tab In Becker (1974) for example, an individual\uc0\u8217 s utility is defined over not only his own consumption, but also the consumption of others (positively in the case of altruism). Collard (1978) distinguishes between this type of altruism, which he calls commodity-related, and altruism in which an individual\uc0\u8217 s utility is defined over his level of consumption and the utility of others (positively in the case of altruism), which he calls utility-related.} \par\pard\ri180\sl480\slmult1{\f8\tab Models of altruism have been influential in explaining economic behavior in many settings, including charitable contributions and volunteer behavior (e.g. Unger, 1991; Smith, Kehoe and Cremer, 1995), social security and other welfare systems (e.g. Coate, 1995), intergenerational bequests and macroeconomic growth (e.g. Rangazas, 1991;}{\f7\b }{\f8 Hori, 1992;}{\f7\b }{\f8 Chakrabarti, Lord and Rangazas, 1993; Strawcyznski, 1994), fertility (e.g. Becker and Barro, 1988; Tamura, 1994), migration (e.g. Tcha, 1995), safety decisions (Jones-Lee, 1992), rent-seeking (Jones, 1996) and behavior in the workplace (e.g. Rotemberg, 1994). Other studies have examined altruism from an evolutionary perspective, either describing evolutionary reasons for altruistic preferences or determining the evolutionary outcomes of societies with heterogenously altruistic individuals (e.g.}{\f7\b }{\f8 Bergstrom and Stark, 1993; Haltiwanger and Waldman, 1993; Samuelson, 1993; Bergstrom, 1995).} \par\pard\ri180\sl480\slmult1{\f8\tab However, recently a number of papers have presented theoretical results which challenge theories of altruism (Warr, 1982; Warr, 1983; Roberts 1984; Sugden, 1985; Bergstrom et al., 1986; Andreoni 1988a; Bernheim and Stark, 1988) as well as empirical data inconsistent with these models. For example, models of pure altruism imply full \uc0\u8220 crowding out\uc0\u8221 of both voluntary contributions and subsidies (Warr, 1982; Roberts, 1984; Bernheim, 1986; Andreoni, 1988a), although there is little evidence of crowding out empirically (Abrams and Schitz, 1978; 1984;}{\f7\b }{\f8 Clotfelter, 1985) or experimentally (Andreoni, 1993). Models of altruism explaining bequests and inter-vivos transfers have similarly found little support in the data (e.g.}{\f7\b }{\f8 Cox, 1987;}{\f7\b }{\f8 Altonji, Hayashi and Kotlikoff, 1992;}{\f7\b }{\f8 Cox and Rank 1992; Hayashi, 1995; Laitner and Juster, 1996;}{\f7\b }{\f8 Wilhelm, 1996) as have models for charitable giving (e.g. Khanna, Posnett and Sandler, 1995).} \par\pard\ri180\sl480\slmult1{\f8\tab Andreoni (1989, 1990) generalized previous models of altruism to incorporate into an individual\uc0\u8217 s utility function not only the consumption (or welfare) of others, but also the \uc0\u8220 warm glow\uc0\u8221 of giving (a related paper, Abel and Warshawsky, 1988, discusses the \uc0\u8220 joy of giving\uc0\u8221 and another,}{\f7\b }{\f8 Feldstein, 1975 models a similar process.). Under this model of impure altruism, an individual cares not only about the consumption of others, but also receives some private goods benefit from their gift }{\f9\i per se.}{\f8 Andreoni (1990) shows that impure altruism implies only partial crowding out, consistent with the empirical results. Models of impure altruism have been used to explain behavior in the supply of charity services by hospitals (Frank and Salkever, 1991) and contributions to public goods and charities in general (Andreoni, 1989, 1990).} \par\pard\ri180\sl480\slmult1{\f8\tab A number of experiments have tested for the existence and malleability of (pure or impure) altruistic preferences. Probably the best-known are experiments in the dictator game (where one individual is given a sum of money to allocate in any way they wish between themselves and another; Camerer and Thaler, 1995 present a review of such experiments). Subjects frequently allocate positive amounts to the other player in the game (Forsythe }{\f9\i et al}{\f8 ., 1994; Hoffman et al., 1994), and the amounts allocated change with the social distance between the players (Hoffman, McCabe and Smith, 1996), the perception of neediness of the recipient (Eckel and Grossman, 1996b) and other institutional factors. Altruism has also been used to explain experimental behavior in prisoners dilemma games (Andreoni and Miller, 1993; Cooper }{\f9\i et al}{\f8 ., 1996), public goods games (Palfrey and Rosenthal, 1988; Andreoni, 1995), and bargaining (Forsythe, Kennan and Sopher, 1991). } \par\pard\ri180\sl480\slmult1{\f8\tab For purposes of our study, we will test the comparative static predictions of models of pure and impure altruism. Under pure altruism, individuals maximize a utility function which includes both their own private consumption and the consumption generated to the group from the public good as below} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 U}{\f17\sub i}{\f8 = U}{\f17\sub i}{\f8 ([\{E}{\f17\sub i}{\f8 - x}{\f17\sub i}{\f8 \} + P}{\f12\uc0\u-4013}{\f17\sub i}{\f8 x}{\f17\sub i}{\f8 ], PN}{\f12\uc0\u-4013}{\f17\sub i}{\f8 x}{\f17\sub i}{\f8 )} \par\pard\ri180\sl480\slmult1{\f8 where U}{\f17\sub i1}{\f8 > 0, U}{\f17\sub i11}{\f8 < 0; U}{\f17\sub i2}{\f8 > 0, U}{\f17\sub i22}{\f8 < 0 (both personal consumption and altruistic consumption are normal goods with decreasing returns)} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 Whenever
0. However, under this assumption }{\f12\uc0\u-3996}{\f8 x}{\f17\sub i}{\f8 */}{\f12\uc0\u-3996}{\f8 x}{\f17\sub j}{\f8 < 0 j\uc0\u8800 iN. This result is akin to crowding out (see Sugden, 1982, p. 346 for a proof).} \par\pard\ri180\sl480\slmult1{\f8\tab Under impure altruism, individuals maximize a utility function which includes the above as well as the amount }{\f9\i they }{\f8 contributed to the public good, as below.} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 U}{\f17\sub i}{\f8 = U}{\f17\sub i}{\f8 ([\{E}{\f17\sub i}{\f8 - x}{\f17\sub i}{\f8 \} + P}{\f12\uc0\u-4013}{\f17\sub i}{\f8 x}{\f17\sub i}{\f8 ], PN}{\f12\uc0\u-4013}{\f17\sub i}{\f8 x}{\f17\sub i}{\f8 , x}{\f17\sub i}{\f8 ) } \par\pard\ri180\sl480\slmult1{\f8 where U}{\f17\sub i1}{\f8 > 0, U}{\f17\sub i11}{\f8 < 0; U}{\f17\sub i2}{\f8 > 0, U}{\f17\sub i22}{\f8 < 0; U}{\f17\sub i3}{\f8 > 0, U}{\f17\sub i33}{\f8 < 0 (personal consumption, altruistic consumption and warm glow consumption are normal goods with decreasing returns)} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 Again, whenever
0. Also under this assumption, }{\f12\uc0\u-3996}{\f8 x}{\f17\sub i}{\f8 */}{\f12\uc0\u-3996}{\f8 x}{\f17\sub j}{\f8 < 0 j\uc0\u8800 iN. This result is akin to partial crowding out; an increase in the amount of the public good provided implies a decrease in an individual\uc0\u8217 s own contribution, although the decrease is smaller than under pure altruism (see Andreoni, 1989, p. 1451 for a proof). Thus our comparative static prediction from both types of theories of altruism (the }{\f9\i altruism hypothesis}{\f8 ) is that there will be a negative relationship between an individual's own contribution and (his beliefs about) the contributions of others in his group.} \par\pard\ri180\sl480\slmult1{\f9\i C. Reciprocity Theories} \par\pard\ri180\sl480\slmult1{\f8\tab A final set of theories of reciprocity assume that individuals reciprocate or match the contributions of others. These theories then go on to generate behavior which involves (1) positive levels of contributions to public goods, but also (2) contributions which are positively related to the contributions of others.} \par\pard\ri180\sl480\slmult1{\f8\tab Sugden (1984) describes a model in which individuals profit-maximize subject to an external constraint; the principle of reciprocity. This principle says that an individual must contribute the minimum of (1) the least any other member of his group is contributing and (2) the level of contribution he would most prefer that every member of the group make (the same as the level of contributions he would make under commitment theories). By assuming this principle as a constraint on behavior, Sugden derives the existence of (multiple) equilibria in settings of both identical and nonidentical players. } \par\pard\ri180\sl480\slmult1{\f8\tab Reciprocal reasoning has been used to explain empirically observed individual behavior in tax evasion (Bordignon, 1993), gift exchange (Solow, 1994; Kranton, 1996), public goods provision (Hollander, 1990), helping in the workplace (Frey, 1993), joint ventures (Kogut, 1989), and labor markets (Akerlof, 1982, 1984; Fehr and Gachter, forthcoming).} \par\pard\ri180\sl480\slmult1{\f8\tab A number of experiments have reported behavior consistent with reciprocity as well. In experimental labor markets, subjects playing the role of firms offer efficiency wages and subjects playing the role of workers respond reciprocally by offering more effort than is individually rational (Fehr, Kirchsteiger and Riedl, 1996; Fehr and Tougareva, 1996; Fehr and Tyran, 1996; Kirchler, Fehr and Evans, 1996; Fehr, Gachter and Kirchsteiger, 1997; Gachter and Falk, 1997; Fehr, Kirchsteiger and Riedl, 1998). A similar result was found in experimental goods markets (Fehr, Kirchsteiger and Riedl, 1993). Reciprocal behavior was also found in experimental bargaining games like the trust game (Berg, Dickhaut and McCabe, 1995; Van Huyck, Battalio and Walters, 1995; Jacobsen and Sadreih, 1996; Abbink, Irlenbusch and Renner, 1997; Guth, Ockenfels and Wendel, 1997; Buchan, Johnson and Croson, 1998) and the common pool resource public goods game (Messick }{\f9\i et al.,}{\f8 1983; Schroeder }{\f9\i et al.,}{\f8 1983; Poppe and Utens, 1986; Wade-Benzoni, Tenbrunsel and Bazerman, 1996). Other bargaining-type experimental games have also exhibited evidence of reciprocity (Bolton, Brandts and Katok, 1996; Bolton, Brandts and Ockenfels, 1997).} \par\pard\ri180\sl480\slmult1{\f8\tab For purposes of our study, we will test the comparative static predictions of models of reciprocity, which predict a significant positive relationship between an individual\uc0\u8217 s contributions to the public good and those of his group. In our notation, under this theory an individual maximizes his personal utility as below} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 U}{\f17\sub i}{\f8 = (E}{\f17\sub i}{\f8 - x}{\f17\sub i}{\f8 ) + P}{\f12\uc0\u-4013}{\f17\sub i}{\f8 x}{\f17\sub i} \par\pard\ri180\sl480\slmult1{\f8\tab subject to\tab x}{\f17\sub i}{\f8 \uc0\u8805 min (x}{\f17\sub i}{\f18\super c}{\f8 , x}{\f17\sub j}{\f8 jN)} \par\pard\ri180\sl480\slmult1{\f8\tab where x}{\f17\sub i}{\f18\super c}{\f8 is the optimal level of contribution under commitment theories} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 Whenever
0. However, under this assumption in equilibria it can be that }{\f12\uc0\u-3996}{\f8 x}{\f17\sub i}{\f8 */}{\f12\uc0\u-3996}{\f8 x}{\f17\sub j}{\f8 > 0 j\uc0\u8800 iN. See Result 4 (p. 780) of Sugden (1984) for a proof. This theory is thus consistent with a positive correlation between one's own contribution and the contribution of other members of the group; this prediction will be our }{\f9\i reciprocity hypothesis}{\f8 .}{\f7\b } \par\pard\ri180\sl480\slmult1{\f7\b\tab}{\f8 It is worth noting that Sugden\uc0\u8217 s model of reciprocity is a model of simultaneous (not sequential) matching of contributions. Players in this game do not wait to see what others have contributed, and then reciprocate their contributions. Instead, everyone makes contributions at the same time, maximizing their self-interest subject to the principle of reciprocity and given their beliefs of others\uc0\u8217 contributions. Thus our test of this theory of reciprocity will (of necessity) be a simultaneous one.} \par\pard\ri180\sl480\slmult1{\f9\i D. Summary} \par\pard\ri180\sl480\slmult1{\f8\tab The experiments reported in this paper allow us to discriminate between the comparative statics of three classes of theories of behavior, all of which have been invoked to explain the voluntary provision of public goods. The first class of theories (commitment rules) predicts no correlation between an individual's contribution and the contributions of others, or his beliefs about them (a similar zero correlation is predicted by traditional theories of full free riding). The second class of theories (pure and impure altruism) predicts a negative correlation. Finally the third class of theories (reciprocity) predicts a positive correlation.} \par\pard\ri180\sl480\slmult1{\f8\tab It should be noted that all these models are models of one-shot behavior. In experiments, however, subjects seldom play equilibria on their first try. Rather, they adjust their behavior and converge toward equilibria. In order to give these equilibria their best chance, the experimental design involves two 10-fold repetitions of a public goods game (consistent with previous experiments, Davis and Holt, 1994; Ledyard, 1995). Since the equilibria described above are equilibria of the stage game, they are also equilibria of the finitely repeated game (Smith, 1990).} \par\pard\ri180\sl480\slmult1{\f8\tab Sections 4, 5 and 6 below present the experiments which distinguish between these different theories.} \par\pard\ri180\sl480\slmult1 \par\pard\ri180\sl480\slmult1{\f7\b 4. Study I: Comparing Contributions with Estimated Contributions of Others} \par\pard\ri180\sl480\slmult1{\f7\b\tab}{\f8 This study tests the comparative static predictions of models of commitment, altruism and reciprocity by comparing an individual\uc0\u8217 s contribution with his belief about the contribution of others. Subjects play a finitely repeated linear public goods game. Before each period, they are asked to estimate the contributions of the other members of their group. The first subsection describes the experimental design and parameters, the second discusses results investigating the comparative statics of behavior and the third addresses the accuracy of subjects\uc0\u8217 beliefs. The fourth subsection concludes.} \par\pard\ri180\sl480\slmult1{\f9\i A. Experimental Design and Parameters} \par\pard\ri180\sl480\slmult1{\f8\tab This experiment was designed to replicate previous experiments in finitely repeated linear public goods games as closely as possible}{\f7\b }{\f8 (Davis and Holt, 1994; Ledyard, 1995). The same structure of game and parameter values as in previous experiments}{\f7\b }{\f8 were used.} \par\pard\ri180\sl480\slmult1{\f8\tab In each period of the game, each subject was endowed with 25 tokens which could be allocated either to a private account, which paid 2\'a2 per token to the individual only, or to a group account (the public good), which paid 1\'a2 per token to each of the four members of the individual\uc0\u8217 s group. Notice that each period of this experiment incorporates a pure public goods problem. Under traditional assumptions of self-interest, regardless of the decisions of the other players, each individual strictly prefers to place all of his tokens in his private account, earning 2\'a2 per token, than in the group account, earning 1\'a2 per token. However the group as a whole earns 4\'a2 when a token is placed in the group account (1\'a2 to each of the four members) but only earns 2\'a2 when the token is placed in a private account. This yields an MPCR of .5, similar to that in previous papers (Davis and Holt, 1994; Ledyard, 1995).} \par\pard\ri180\sl480\slmult1{\f8\tab In a departure from previous experiments, each period of the game was divided into two stages; the \uc0\u8220 guessing\uc0\u8221 and the \uc0\u8220 decision\uc0\u8221 stage. In the guessing stage, subjects estimated the total number of tokens the other three members of their group would contribute to the group account in the upcoming decision stage. They were compensated for accurate estimates. In the decision stage, subjects made their personal and private contribution decisions, as in previous experiments.} \par\pard\ri180\sl480\slmult1{\f8\tab At the end of each period, subjects were reminded of their own estimate, told the true aggregate contribution of the other three members of group, the total group contribution and their earnings from both the estimation stage and the contribution stage.} \par\pard\ri180\sl480\slmult1{\f8\tab Subjects played two identical ten-period linear public goods games. Subjects played the first game, and then were told there was enough time to play a second (as in Andreoni, 1988b). The second game was always identical to the first.} \par\pard\ri180\sl480\slmult1{\f7\b\tab}{\f8 Twenty-four subjects arranged in six groups of four participated in this experiment. Subjects were undergraduate students at the University of Arizona summer session. They were paid a five dollar show-up fee along with their earnings in the experiment. Average earnings were $14.69, plus the $5 fee, for less than an hour of experimental time. The entire experiment was computerized; instructions were given through the computer screen, subjects entered their contributions via the keyboard and, at the end of each period, feedback about the outcome was displayed on the screen. Subjects could also access a \uc0\u8220 history\uc0\u8221 of past outcomes of their group at any time.} \par\pard\ri180\sl480\slmult1{\f8\tab In the following subsections we directly test the comparative statics of commitment, altruism and reciprocity theories by investigating the relationship between an individual\uc0\u8217 s}{\f9\i }{\f8 beliefs of what others will contribute and his own contributions. Commitment theories predict a zero correlation, altruism theories a negative correlation and reciprocity theories a positive correlation.} \par\pard\ri180\sl480\slmult1{\f9\i B. Results: Testing Comparative Statics} \par\pard\ri180\sl480\slmult1{\f19\ul 1. Overall} \par\pard\ri180\sl480\slmult1{\f8\tab A random effects regression as below compares an individual's contribution with his belief of what the rest of his group will contribute.} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 CONT}{\f17\sub it}{\f8 = }{\f12\uc0\u-3999}{\f17\sub 0}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 1}{\f8 GUESS}{\f17\sub t}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 2}{\f8 PERIOD + }{\f12\uc0\u-4013}{\f17\sub i\uc0\u8800 1}{\f12\uc0\u-3998}{\f17\sub i}{\f8 IND + }{\f12\uc0\u-3995}{\f17\sub i} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 The dependent variable is an individual i\uc0\u8217 s contribution to the public good in period t. Independent variables are the individual's GUESS of what the other three members of his group will contribute in this same period t, the PERIOD number, and an indicator variable for each individual (IND). A random effects regression not only allows for individual-specific intercepts (from the indicator variables) but also individual-specific error terms. For a complete discussion of random effects regression see Greene, 1990. Results of the regression are reported in Table 1, parameter estimates for individual dummies are suppressed for ease of presentation.} \par\pard\ri180{\f4 _______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 1 about here} \par\pard\ri180{\f4 _______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab This regression reports a significant positive relationship between a subject's guess of what the other three members of his group will contribute and his own contribution. This result strongly supports reciprocity theories over theories of commitment (which predicted a zero correlation) and altruism (which predicted a negative correlation).} \par\pard\ri180\sl480\slmult1{\f8\tab The intercept in this regression was also positive, consistent with previous experimental results that subjects make positive contributions in similar games (Davis and Holt, 1994; Ledyard, 1995). The coefficient on period is not significant.} \par\pard\ri180\sl480\slmult1{\f8\tab Identical regressions using only the data from the first game (the first 10 periods) or the second game (the second 10 periods) yield similar results, reported in Table 2.} \par\pard\ri180{\f4 _______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 2 about here} \par\pard\ri180{\f4 _______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab The intercept and GUESS coefficient is similarly positive and significant in each individual game. The PERIOD coefficient is significantly negative in the first game, suggesting declining contributions over time, also observed in previous experiments, (Davis and Holt, 1994; Ledyard, 1995). However, by the second game contributions appear to have stabilized and no decrease is observed.} \par\pard\ri180\sl480\slmult1{\f8\tab While these analyses examine the correlation between contributions and beliefs over time, alternatively, we can investigate the between-subject correlation between contributions and beliefs in only the first or only the last periods of the game. To do this, we estimate the following random effects OLS regression} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 CONT}{\f17\sub it}{\f8 = }{\f12\uc0\u-3999}{\f17\sub 0}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 1}{\f8 GUESS}{\f17\sub t}{\f8 + }{\f12\uc0\u-4013}{\f17\sub i\uc0\u8800 1}{\f12\uc0\u-3998}{\f17\sub i}{\f8 IND + }{\f12\uc0\u-3995}{\f17\sub i} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 separately for only period 1 and only period 10. Table 3 provides results from these regressions.} \par\pard\ri180{\f4 _______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 3 about here} \par\pard\ri180{\f4 _______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab Again, consistent with the reciprocity model, we find a significantly positive relationship between an individual\uc0\u8217 s contributions and his estimates of others\uc0\u8217 contributions. In addition, we observe positive (but highly variable) contributions in period 1 of the game, and significantly lower contributions in period 10 of the games, consistent with previous experiments in this area which document decreasing contributions over time (Davis and Holt, 1994; Ledyard, 1995).} \par\pard\ri180\sl480\slmult1{\f19\ul 2. Individual Characterizations} \par\pard\ri180\sl480\slmult1{\f8\tab A second type of analysis characterizes the behavior of individual subjects in the experiment. For each of the 24 subjects, we calculate the correlation between their contribution and their belief of the contribution of others in their group. In this way, we can identify individual subjects whose behavior is consistent with comparative statics of commitment, altruism or reciprocity models. Table 4 depicts the results from this analysis. } \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 4 about here} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab Twenty-two out of 24 subjects (almost 92%) exhibit a positive correlation between their own contribution and their estimates of the contributions of others, consistent with models of reciprocity. Only two subjects exhibit a negative correlation, consistent with models of altruism, and none a zero correlation, consistent with models of commitment.} \par\pard\ri180\sl480\slmult1{\f8\tab These results represent a statistically significant difference from random behavior. A chi-squared test comparing the actual categorization of subjects against a null hypothesis of equal probability of all three types, rejects the null at p<.01.} \par\pard\ri180\sl480\slmult1{\f9\i C. Results: Estimate Accuracy} \par\pard\ri180\sl480\slmult1{\f8\tab In this experiment, each subject estimated what others in his group would contribute. One important question involves the accuracy of these estimates.} \par\pard\ri180\sl480\slmult1{\f8\tab Figure 1 shows the average absolute estimation error made by each group in each period of the game. This error is calculated by computing the }{\f9\i absolute}{\f8 error of each subject in each period (the distance between their guess and the other three subjects\uc0\u8217 actual contributions) and averaging within each group. If all subjects were extremely bad guessers this average absolute estimation error could be as high as 25. Instead subjects appear to be fairly accurate in their estimations of others\uc0\u8217 behavior.} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180{\f4 Insert Figure 1 about here} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab We can define an individual subject\uc0\u8217 s estimation error as the difference between that subject\uc0\u8217 s estimate and the actual contributions of the other members of his group. Over all ten rounds of the first game, only eight out of 24 subjects exhibited any significantly positive levels of error (overoptimism). In the second game, no subjects exhibited significant levels of error. Throughout the experiment, most subjects provided unbiased guesses of what their counterparts in the public goods game will do.} \par\pard\ri180\sl480\slmult1{\f9\i D. Conclusion} \par\pard\ri180\sl480\slmult1{\f8\tab This study was designed to provide data which could distinguish between models of commitment, altruism and reciprocity by comparing their comparative static predictions. In particular, we compare an individual\uc0\u8217 s contribution in a public goods game with his beliefs about the contributions of others. Results from a random effects regression demonstrate a significant positive relationship, consistent with models of reciprocity and inconsistent with models of commitment or altruism. In addition, an analysis at the level of individual subject is run. Almost 92% of the subjects (22 out of 24) demonstrate a positive correlation between their own contributions and their beliefs of others contributions, consistent with reciprocity models. The remaining 8% exhibit a negative correlation, consistent with models of altruism and none exhibit a zero correlation, consistent with models of commitment and of traditional self-interest models.} \par\pard\ri180\sl480\slmult1{\f8\tab In addition, we provide an analysis of the accuracy of subjects\uc0\u8217 beliefs in this setting. We find that subjects\uc0\u8217 beliefs are quite accurate. In the first game, only eight subjects out of 24 had significantly positive levels of error; in the second game no subjects did.} \par\pard\ri180\sl480\slmult1{\f8\tab Although these results appear encouraging for reciprocity models, a few questions remain. First, it may be that asking subjects for their estimates of others\uc0\u8217 actions leads them to think reciprocally where they wouldn\uc0\u8217 t otherwise (an elicitation hypothesis). Second, it may be that the repeated game nature of this experiment is yielding the positive correlation and not reciprocity }{\f9\i per se}{\f8 (a reputation hypothesis).} \par\pard\ri180\sl480\slmult1{\f8\tab To answer these questions and test the robustness of our results, two more experiments were run and the comparative static predictions of our models re-analyzed. Neither of the experiments involved the elicitation of beliefs of others\uc0\u8217 actions. Instead, we compare an individual\uc0\u8217 s contribution with the actual contribution of the other members of his group. Since in this experiment, subjects\uc0\u8217 elicited beliefs were quite accurate, we claim the comparative static predictions from the models will transfer to this new context. The next section describes the new experiments and their results.} \par\pard\ri180\sl480\slmult1 \par\pard\ri180\sl480\slmult1{\f7\b 5. Study II: Comparing Contributions with Actual Contributions of Others} \par\pard\ri180\sl480\slmult1{\f9\i A.. Testing the Elicitation Hypothesis} \par\pard\ri180\sl480\slmult1{\f8\tab Our first question involves the extent to which asking subjects to estimate the contributions of others leads them to play reciprocally where they wouldn\uc0\u8217 t otherwise. To test this hypothesis, we ran a new experiment, identical to the first but excluding the estimation stage. Twenty-four subjects, different from the previous subjects but from the same subject pool, participated in this experiment, arranged in six groups of four. Average earnings for this experiment were $13.91 (plus the $5 show-up fee) for less than one hour of experimental time.} \par\pard\ri180\sl480\slmult1{\f8\tab Here, we are interested in the correlation between subjects\uc0\u8217 contribution and the ACTUAL total contribution of others in this same period (unfortunately, we can\uc0\u8217 t compare an individual\uc0\u8217 s contribution with his belief about the contributions of others when those beliefs are not elicited). A zero correlation is predicted by commitment theories, a negative correlation by altruism theories and a positive correlation by reciprocity theories. With this data, we estimate the random effects OLS regression} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 CONT}{\f17\sub it}{\f8 = }{\f12\uc0\u-3999}{\f17\sub 0}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 1}{\f8 ACTUAL}{\f17\sub t}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 2}{\f8 PERIOD + }{\f12\uc0\u-4013}{\f17\sub i\uc0\u8800 1}{\f12\uc0\u-3998}{\f17\sub i}{\f8 IND + }{\f12\uc0\u-3995}{\f17\sub i} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 for both the previous experiment and this one. Results from these regressions are reported in Table 5, below.} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 5 about here} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab Consistent with the reciprocity hypothesis, a significant positive relationship is found between an individual\uc0\u8217 s contribution and the actual contribution of others in both the previous experiment (Guess) and this experiment (No Guess). First, focusing on the previous experiment, this relationship is a bit less strong than the relationship between an individual\uc0\u8217 s contribution and estimate of others\uc0\u8217 contributions (where the coefficient was .202 rather than .164). The relationship appears somewhat weaker in this experiment (.077 versus .164), nonetheless it is still significant at the 1% level.} \par\pard\ri180\sl480\slmult1{\f8\tab Interestingly, subjects contributed on average significantly less in the previous experiment (when they were asked to estimate the contributions of others) than in this one, as can be seen by the difference in intercept between the two regressions. This difference is explored in more detail in a related paper, Croson (1997). Finally, both regressions show a significant decrease in contributions over the course of the games, consistent with evidence from previous experiments (Davis and Holt, 1994; Ledyard, 1995).} \par\pard\ri180\sl480\slmult1{\f8\tab In a parallel to our first analysis, we can also compare within an individual, their correlation between their contributions and the actual contributions of others in their group. Table 6 provides those results for the previous experiment and this experiment.} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 6 about here} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab In the previous experiment (Guess), 21 out of 24 subjects (87.5%) exhibited a positive correlation between their own contribution and the actual contributions of others in their group, consistent with reciprocity theories. Only 3 subjects exhibited a negative correlation, consistent with altruism theories and no subjects exhibited a zero correlation. The pattern in this experiment (No Guess) is similar, 19 out of 24 subject exhibited a positive correlation (79%), only 5 a negative correlation and none exhibited a zero correlation. These results represent a statistically significant difference from random behavior. A chi-squared test comparing the actual categorization of subjects against a null hypothesis of equal probability of all three types, rejects the null at p<.01, for each experiment independently as well as for both of them together.} \par\pard\ri180\sl480\slmult1{\f8\tab While there are some differences in the }{\f9\i level }{\f8 of contributions between games in which estimates of others\uc0\u8217 contributions are elicited and where they are not, results from this subsection demonstrate that the }{\f9\i comparative statics}{\f8 of reciprocity theories remain most consistent with the data, even when beliefs are not elicited. This allows us to reject the elicitation hypothesis.} \par\pard\ri180\sl480\slmult1{\f9\i B. Testing the Reputation Hypothesis} \par\pard\ri180\sl480\slmult1{\f8\tab A further concern is that the positive correlation observed is arising from some sort of reputation development rather than from reciprocity. In both our original experiment and the second experiment presented above, subjects were formed into groups which remained constant for the duration of the experiment. This matching procedure is by far the most common one used in previous experiments (Davis and Holt, 1994; Ledyard, 1995). However, in this setting it may be another cause of our results. For example, if all subjects were playing in a way consistent with Kreps}{\f9\i et al.}{\f8 (1986), and all subjects believed all subjects were playing in this way, we might observe a positive correlation similar to the one we observed, but for reputational rather than reciprocal reasons. The experiment reported in this subsection was designed to test this alternative explanation.} \par\pard\ri180\sl480\slmult1{\f8\tab A different 24 subjects participated in this experiment (Strangers). The experiment was run in two separate sessions of 12 subjects each. Subjects played the same game as in the previous experiments. After each period of the game, however, subjects were randomly reassigned to new groups of four (as in Andreoni, 1988b; Croson, 1996; van Winden and Keser, 1997). Thus it was extremely unlikely a subject would play with the same group of three other people more than once during the session. This type of matching scheme has been demonstrated to reduce reputation effects (for an explanation of why, see Andreoni, 1988b). Average earnings for this experiment were $11.83 (plus the $5 show-up fee) for less than one hour of experimental time.} \par\pard\ri180\sl480\slmult1{\f8\tab Here, we are again interested in the correlation between subjects\uc0\u8217 contribution and the ACTUAL total contribution of others with whom he is matched in this same period. If the previously-observed positive correlation is being caused by reputational issues (the reputation hypothesis) we should observe a zero correlation in this experiment. Thus, with this data, we estimate the random effects regression} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 CONT}{\f17\sub it}{\f8 = }{\f12\uc0\u-3999}{\f17\sub 0}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 1}{\f8 ACTUAL}{\f17\sub t}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 2}{\f8 PERIOD + }{\f12\uc0\u-4013}{\f17\sub i\uc0\u8800 1}{\f12\uc0\u-3998}{\f17\sub i}{\f8 IND + }{\f12\uc0\u-3995}{\f17\sub i} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 for this experiment. Results from these regressions are reported in Table 7, below.} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 7 about here} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab Consistent with the reciprocity hypothesis, a significant positive relationship is found between an individual\uc0\u8217 s contribution and the actual contribution of others in this experiment as well. Although the relationship appears somewhat weaker in this experiment than in previous ones (suggesting some reputation formation may be going on), nonetheless it is still significant at the 1% level, supporting theories of reciprocity over those of commitment or altruism.} \par\pard\ri180\sl480\slmult1{\f8\tab In similar analysis to that above, we also compare, within an individual, the correlation between their contributions and the actual contributions of others in their group. Table 8 provides those results for this experiment.} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 8 about here} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab As in previous experiments, most of the subjects exhibited positive correlations between their own contributions and contributions of others, consistent with reciprocity theories (almost 71%). Only three out of 24 subjects exhibited negative correlations, consistent with altruism theories. In contrast to previous experiments, however, four subjects exhibited zero correlations between their own contributions and the contributions of others. Closer inspection reveals that these correlations were generated by four subjects who fully free rode (contributed zero) throughout the entire experiment. This result of more free riding and lower contributions in strangers experiments than among stable groups is consistent with previous research (Croson, 1996; van Winden and Keser, 1997). } \par\pard\ri180\sl480\slmult1{\f8\tab These results represent a statistically significant difference from random behavior. A chi-squared test comparing the actual categorization of subjects against a null hypothesis of equal probability of all three types, rejects the null at p<.01.} \par\pard\ri180\sl480\slmult1{\f8\tab While there are more free riders and lower contributions in this experiment than in previous ones, results are still supportive of the comparative statics of reciprocity theories over those of commitment or altruism. A significant and positive relationship is found between an individual\uc0\u8217 s contribution and the contribution of others in his group.} \par\pard\ri180\sl480\slmult1{\f9\i C. Conclusion} \par\pard\ri180\sl480\slmult1{\f8\tab In this section we presented the results of two further experiments, which test the robustness of our previous result. The first demonstrates a positive relationship between an individual\uc0\u8217 s contribution and the contributions of others even when beliefs are not elicited. The second demonstrates a similarly positive relationship even when reputational concerns are severely reduced. We conclude that reciprocal concerns in this setting are robust. Having demonstrated support for theories of reciprocity in three different settings, we now turn to Study III, which provides a characterization of the }{\f9\i type}{\f8 of reciprocity individuals exhibit. } \par\pard\ri180\sl480\slmult1 \par\pard\ri180\sl480\slmult1{\f7\b 6. Study III: Types of Reciprocity} \par\pard\ri180\sl480\slmult1{\f8\tab In Sugden\uc0\u8217 s (1984) model of reciprocity, he suggests that actors will match the }{\f9\i minimum}{\f8 contribution of others. In contrast, however, we can imagine different }{\f9\i types }{\f8 of reciprocity in which subjects try to match the average contribution of others, or possibly even the maximum. The experiment reported in this study distinguishes between these different specifications of reciprocity.} \par\pard\ri180\sl480\slmult1{\f9\i A. Experimental Design and Parameters} \par\pard\ri180\sl480\slmult1{\f8\tab Twenty-four subjects, distinct from previous participants but from the same subject pool, participated in this experiment. Subjects were arranged into groups of four and played two ten-round games retaining the same groups. All parameter values were the same as in the previous experiments, and no elicitation of beliefs was made.} \par\pard\ri180\sl480\slmult1{\f8\tab In contrast to the previous experiments, however, after each period subjects were informed not only of the aggregate contribution of the other three members of their group, but also of the }{\f9\i individual }{\f8 contributions of the other three members of the group (as in Sell and Wilson, 1991; Croson, forthcoming). Thus subjects could attempt to match either the maximum, the minimum or the middle contribution. As before, subjects took home their earnings from the experiment (average $14.03) plus their $5 show-up-fee.} \par\pard\ri180\sl480\slmult1{\f9\i B. Results: Testing Comparative Statics} \par\pard\ri180\sl480\slmult1{\f8\tab In this experiment we again observe a positive relationship between an individual\uc0\u8217 s own contribution and the actual contribution of others in their group. With this data, we again estimate the random effects OLS regression} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 CONT}{\f17\sub it}{\f8 = }{\f12\uc0\u-3999}{\f17\sub 0}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 1}{\f8 ACTUAL}{\f17\sub t}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 2}{\f8 PERIOD + }{\f12\uc0\u-4013}{\f17\sub i\uc0\u8800 1}{\f12\uc0\u-3998}{\f17\sub i}{\f8 IND + }{\f12\uc0\u-3995}{\f17\sub i} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 Results from this regression are shown in Table 9, below.} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 9 about here} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab As before, we observe a significantly positive relationship between one\uc0\u8217 s own contribution and the actual contributions of others in one\uc0\u8217 s group. The intercept is significantly positive, and the period variable significantly negative, consistent with results from previous experiments (Davis and Holt, 1994; Ledyard, 1995).} \par\pard\ri180\sl480\slmult1{\f8\tab A parallel analysis as above involving correlations at the individual level also yields similar results, as shown in Table 10.} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 10 about here} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab Almost 88% (21 subjects out of 24) exhibit a positive correlation, consistent with theories of reciprocity. Only two subjects out of 24 (8.33%), exhibit a negative correlation, consistent with theories of altruism. One subject\uc0\u8217 s behavior yields a zero correlation, closer inspection shows that subject contributes nothing throughout the entire experiment. Thus we classify him as a free rider. These results represent a statistically significant difference from random behavior. A chi-squared test comparing the actual categorization of subjects against a null hypothesis of equal probability of all three types, rejects the null at p<.01.} \par\pard\ri180\sl480\slmult1{\f8\tab This experiment demonstrates similar evidence in favor of theories of reciprocity as previous ones. In addition, however, we can use the data to compare between different types of reciprocity. This is done in the next subsection.} \par\pard\ri180\sl480\slmult1{\f9\i C. Different Types of Reciprocity} \par\pard\ri180\sl480\slmult1{\f8\tab This data allows us to compare three different types of reciprocity: maximum, minimum and middle. In maximum reciprocity, subjects would attempt to match the maximum contribution of the other three members of their group. In minimum reciprocity, subjects would attempt to match the minimum contribution of the other three members of their group. Finally, in middle reciprocity, subjects would attempt to match the contributions of the middle contributor of the other three members of their group.} \par\pard\ri180\sl480\slmult1{\f8\tab Our goal is to determine which of these three models of reciprocity best fits the data. That is, which of the minimum, maximum or middle contribution of an individual\uc0\u8217 s partners better predicts an individual\uc0\u8217 s own contribution. We estimate four random effect OLS regressions, as below} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 CONT}{\f17\sub it}{\f8 = }{\f12\uc0\u-3999}{\f17\sub 0}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 1}{\f8 MIN}{\f17\sub t}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 2}{\f8 PERIOD + }{\f12\uc0\u-4013}{\f17\sub i\uc0\u8800 1}{\f12\uc0\u-3998}{\f17\sub i}{\f8 IND + }{\f12\uc0\u-3995}{\f17\sub i} \par\pard\ri180\sl480\slmult1{\f8 CONT}{\f17\sub it}{\f8 = }{\f12\uc0\u-3999}{\f17\sub 0}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 1}{\f8 MAX}{\f17\sub t}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 2}{\f8 PERIOD + }{\f12\uc0\u-4013}{\f17\sub i\uc0\u8800 1}{\f12\uc0\u-3998}{\f17\sub i}{\f8 IND + }{\f12\uc0\u-3995}{\f17\sub i} \par\pard\ri180\sl480\slmult1{\f8 CONT}{\f17\sub it}{\f8 = }{\f12\uc0\u-3999}{\f17\sub 0}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 1}{\f8 MID}{\f17\sub t}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 2}{\f8 PERIOD + }{\f12\uc0\u-4013}{\f17\sub i\uc0\u8800 1}{\f12\uc0\u-3998}{\f17\sub i}{\f8 IND + }{\f12\uc0\u-3995}{\f17\sub i} \par\pard\ri180\sl480\slmult1{\f8 CONT}{\f17\sub it}{\f8 = }{\f12\uc0\u-3999}{\f17\sub 0}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 1}{\f8 MIN}{\f17\sub t}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 2}{\f8 MAX}{\f17\sub t}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 3}{\f8 MID}{\f17\sub t}{\f8 + }{\f12\uc0\u-3999}{\f17\sub 4}{\f8 PERIOD + }{\f12\uc0\u-4013}{\f17\sub i\uc0\u8800 1}{\f12\uc0\u-3998}{\f17\sub i}{\f8 IND + }{\f12\uc0\u-3995}{\f17\sub i} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8 Results from these regressions are shown in Table 11.} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 11 about here} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab Evidence from these regressions suggests that middle reciprocity is a better predictor than either minimum or maximum reciprocity. First, in the individual regressions (1), (2) and (3), the t-statistic is higher for the middle contribution than for either of the others. In addition, in the regression which includes all measures (4), only middle is significantly different than zero. This suggests that the middle contribution of the others in a subject\uc0\u8217 s group is a better predictor of that subject\uc0\u8217 s own contribution than either the maximum or the minimum. In all of these regressions we observe a similar result as above of significantly positive contributions (positive intercept coefficient) which decrease over time (negative PERIOD coefficient) as in previous studies (Davis and Holt, 1994; Ledyard, 1995).} \par\pard\ri180\sl480\slmult1{\f8\tab Another way to demonstrate this relationship is through a standardized regression. Here, the same regression equations are run as above, except the independent measures (MIN, MAX and MID) are standardized to be distributed normally with mean zero and variance one. In these regressions, the absolute size of the coefficients can be compared directly. Results from the standardized random effects OLS regressions are shown in Table 12.} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180{\f4 Insert Table 12 about here} \par\pard\ri180{\f4 ______________________________________________________________________} \par\pard\ri180 \par\pard\ri180\sl480\slmult1{\f8\tab Again, we see that the middle contribution of the other players is a better predictor of a subject\uc0\u8217 s own contribution than either the maximum or the minimum. The standardized coefficient on MID is higher than either of the other two, and in the final regression (4), only MID remains significantly different than zero. And again we observe significantly positive contributions (positive intercept coefficient) which decrease over time (negative PERIOD coefficient) as in previous studies (Davis and Holt, 1994; Ledyard, 1995).} \par\pard\ri180\sl480\slmult1{\f9\i D. Conclusion} \par\pard\ri180\sl480\slmult1{\f8\tab This experiment sheds light on exactly what subjects in this experiment might be trying to reciprocate. First, our results are consistent with the comparative statics of reciprocity theories in yet another setting, this time where subjects are given information about the full distribution of their group\uc0\u8217 s contributions, not just the total. Then, we test whether the minimum, maximum or middle contribution of the other three players is a better predictor of a subject\uc0\u8217 s own contribution. We find significant evidence for middle reciprocity, suggesting that subjects try to match the median or average contributions of others, rather than the minimum (as suggested by Sugden\uc0\u8217 s theory of reciprocity) or the maximum.} \par\pard\ri180\sl480\slmult1 \par\pard\ri180\sl480\slmult1{\f7\b 7. Discussion and Conclusion} \par\pard\ri180\sl480\slmult1{\f8\tab The experiments reported in this paper tested comparative statics predictions of three models consistent with observations of voluntary public goods provision; commitment, altruism and reciprocity. The results support the reciprocity model in which individual contributions are positively related to the contributions of others, or to their beliefs about those contributions.} \par\pard\ri180\sl480\slmult1{\f8\tab Reciprocal behavior is also supported by anecdotal evidence. Charities eliciting contributions often suggest a particular level as the "standard" gift or report the size of their "average" contribution. Presumably this influences individual's beliefs of what others are giving, thus causing them to give more.} \par\pard\ri180\sl480\slmult1{\f8\tab Even the very wealthy seem to exhibit reciprocal behavior, in this example from }{\f9\i Forbes Magazine}{\f8 , "Seattle's Lakeside Upper School counts ... Bill Gates among its alumni. Rumor has it a fundraiser for the high school put the bite on Gates, who asked: \uc0\u8216 How much is everyone else giving?\uc0\u8217 About $75 he was told. \uc0\u8216 So put me down for $75,\uc0\u8217 said Gates.\uc0\u8221 (January 22, 1996, p. 18).} \par\pard\ri180\sl480\slmult1{\f8\tab While middle-matching behavior like this is consistent with the reciprocity principle, it may be adaptively rational as well. Societies whose members follow this principle are more likely to be able to supply public goods than societies whose members practice self-interest utility maximization. One can also imagine an individually rational reason to behave reciprocally. If the quality or reliability of charitable groups are not known, individual contributors may use the contributions of others as a signal for how much they should contribute themselves (as in Vesterlund, 1998).} \par\pard\ri180\sl480\slmult1{\f8\tab This study examines the factors that motivate individuals to make voluntary contributions in social dilemma situations. In particular, it finds support for reciprocity theories over commitment theories, altruistic theories and traditional free-riding theories. We find a significant and positive relationship between an individual\uc0\u8217 s contribution and his belief about the contributions of others in his group, as well as between an individual\uc0\u8217 s contribution and the actual contributions of the others in his group. These results suggest that players act as though part of their objective is to match the contributions of other members.} \par\pard\fi-360\li360\ri180\sl480\slmult1{\f20\b\page 8. References} \par\pard \par\pard{\f21 Abbink, Klaus, Bernd Irlenbusch and Elke Renner (1997). \uc0\u8220 The Moonlighting Game: An Experimental Study on Reciprocity and Retribution.\uc0\u8221 Discussion Paper B-415, University of Bonn.} \par\pard \par\pard{\f21 Abel, Andrew and Mark Warshawsky (1988). \uc0\u8220 Specification of the Joy of Giving: Insights from Altruism.\uc0\u8221 }{\f22\i Review of Economics and Statistics}{\f21 , vol 70, pp. 145-149.} \par\pard \par\pard{\f21 Abrams, Burton and Mark Schitz (1978). \uc0\u8220 The Crowding Out Effect of Governmental Transfers on Private Charitable Contributions.\uc0\u8221 }{\f22\i Public Choice}{\f21 , vol 33, pp. 29-39.} \par\pard \par\pard{\f21 Abrams, Burton and Mark Schitz (1984). \uc0\u8220 The Crowding Out Effect of Governmental Transfers on Private Charitable Contributions: Cross Sectional Evidence.\uc0\u8221 }{\f22\i National Tax Journal}{\f21 , vol 37, pp. 563-568.} \par\pard \par\pard{\f21 Akerlof, George (1982). \uc0\u8220 Labor Contracts as Partial Gift Exchange.\uc0\u8221 }{\f22\i Quarterly Journal of Economics}{\f21 , vol 97, pp. 543-569.} \par\pard \par\pard{\f21 Akerlof, George (1984). \uc0\u8220 Gift Exchange and Efficiency-Wage Theory: Four Views.\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 74, pp. 79-83.} \par\pard \par\pard{\f21 Altonji, Joseph, Fumio Hayashi and Laurence Kotlikoff (1992). Is the Extended Family Altruistically Linked? Direct Tests Using Micro Data.\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 82, pp. 1177-1198.} \par\pard \par\pard\tx1260{\f23 Andreoni, James (1988a). \uc0\u8220 Privately Provided Public Goods in a Large Economy: The Limits of Altruism.\uc0\u8221 }{\f24\i Journal of Public Economics}{\f23 , vol 35, pp. 57-73.} \par\pard\tx1260 \par\pard\tx1260{\f23 Andreoni, James (1988b). \uc0\u8220 Why Free Ride? Strategies and Learning in Public Goods Experiments.\uc0\u8221 }{\f24\i Journal of Public Economics}{\f23 , vol 37, pp. 291-304.} \par\pard\tx1260 \par\pard\tx1260{\f23 Andreoni, James (1989). "Giving with Impure Altruism: Applications to Charity and Ricardian Equivalence." }{\f24\i Journal of Political Economy}{\f23 , vol 97, pp. 1447-1458.} \par\pard\tx1260 \par\pard\tx1260{\f23 Andreoni, James (1990). \uc0\u8220 Impure Altruism and Donations to Public Goods: A Theory of Warm-Glow Giving.\uc0\u8221 }{\f24\i Economic Journal,}{\f23 vol 100, pp. 464-477.} \par\pard\tx1260 \par\pard\tx1260{\f23 Andreoni, James (1993). \uc0\u8220 An Experimental Test of the Public Goods Crowding-Out Hypothesis.\uc0\u8221 }{\f24\i American Economic Review}{\f23 , vol 83, pp. 1317-1327.} \par\pard\tx1260 \par\pard\tx1260{\f23 Andreoni, James (1995). \uc0\u8220 Cooperation in Public-Goods Experiments: Kindness or Confusion?\uc0\u8221 }{\f24\i American Economic Review}{\f23 , vol 85, pp. 891-904.} \par\pard\tx1260 \par\pard\tx1260{\f23 Andreoni, James and John Miller (1993). \uc0\u8220 Rational Cooperation in the Finitely Repeated Prisoner\uc0\u8217 s Dilemma: Experimental Evidence.\uc0\u8221 }{\f24\i Economic Journal}{\f23 , vol 103, pp. 570-585.} \par\pard\tx1260 \par\pard\tx1260{\f23 Andreoni, James and Lise Vesterlund (1997). \uc0\u8220 Which is the Fair Sex? Gender Differences in Altruism.\uc0\u8221 Working Paper, University of Wisconsin} \par\pard\tx1260 \par\pard{\f21 Arrow, Kenneth (1975 ). \uc0\u8220 Gifts and Exchanges.\uc0\u8221 in }{\f22\i Altruism, Morality and Economic Theory}{\f21 , E.S. Phelps, ed. New York, NY: Russell Sage Foundation.} \par\pard\tx1260 \par\pard\tx1260{\f23 Asch, Peter, Gary Gigliotti and James Polito (1993). \uc0\u8220 Free Riding with Discrete and Continuous Public Goods: Some Experimental Evidence.\uc0\u8221 }{\f24\i Public Choice}{\f23 , vol 77, pp. 293-305.} \par\pard\tx1260 \par\pard{\f21 Baron, Jon and Mark Spranca (1997). \uc0\u8220 Protected Values.\uc0\u8221 }{\f22\i Organizational Behavior and Human Decision Processes}{\f21 , vol 70, pp. 1-16.} \par\pard \par\pard{\f21 Baldry, Jonathan (1987). \uc0\u8220 Income Tax Evasion and the Tax Schedule.\uc0\u8221 }{\f22\i Public Finance}{\f21 , vol 42, pp. 357-383.} \par\pard \par\pard\tx1260{\f23 Becker, Gary (1974). \uc0\u8220 A Theory of Social Interaction.\uc0\u8221 }{\f24\i Journal of Political Economy}{\f23 , vol 82, pp. 1063-1093.} \par\pard \par\pard{\f21 Becker, Gary and Robert Barro (1988). \uc0\u8220 A Reformulation of the Economic Theory of Fertility.\uc0\u8221 }{\f22\i Quarterly Journal of Economics}{\f21 , vol 103, pp. 1-25.} \par\pard \par\pard{\f21 Berg, Joyce, John Dickhaut and Kevin McCabe (1995). \uc0\u8220 Trust, Reciprocity and Social History.\uc0\u8221 }{\f22\i Games and Economic Behavior}{\f21 , vol 10, pp. 122-142.} \par\pard \par\pard{\f21 Bergstrom, Theodore (1995). \uc0\u8220 On the Evolution of Altruistic Ethical Rules for Siblings.\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 85, pp. 58-81.} \par\pard \par\pard{\f21 Bergstrom, Theodore and Oded Stark (1993). \uc0\u8220 How Altruism Can Prevail in an Evolutionary Environment.\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 83, pp. 149-155.} \par\pard \par\pard{\f21 Bernheim, Douglas (1986). \uc0\u8220 On the Voluntary and Involuntary Provision of Public Goods.\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 76, ppp. 789-793.} \par\pard \par\pard{\f21 Bernheim, Douglas and Oded Stark (1988). \uc0\u8220 Altruism within the Family Reconsidered: Do Nice Guys Finish Last?\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 78, pp. 1034-1045.} \par\pard \par\pard{\f21 Bolton, Gary, Jordi Brandts and Elena Katok (1996). \uc0\u8220 A Simple Test of Explanations for Contributions in Social Dilemmas.\uc0\u8221 Working Paper, Pennsylvania State University.} \par\pard \par\pard{\f21 Bolton, Gary, Jordi Brandts and Axel Ockenfels (1997). Measuring Motivations for the Reciprocal Responses Observed in a Simple Dilemma Game.\uc0\u8221 Working Paper, Pennsylvania State University.} \par\pard \par\pard{\f21 Bolton, Gary and Elena Katok (1995). \uc0\u8220 An Experimental Test for Gender Differences in Beneficent Behavior.\uc0\u8221 }{\f22\i Economics Letters}{\f21 , vol 48, pp. 287-292.} \par\pard \par\pard{\f21 Bordignon, Massimo (1990). \uc0\u8220 Was Kant Right? Voluntary Provision of Public Goods under the Principle of Unconditional Commitment.\uc0\u8221 }{\f22\i Economic Notes}{\f21 , vol 0, pp. 342-372.} \par\pard \par\pard{\f21 Bordignon, Massimo (1993). \uc0\u8220 A Fairness Approach to Income-Tax Evasion.\uc0\u8221 }{\f22\i Journal of Public Economics}{\f21 , vol 52, pp. 345-362.} \par\pard \par\pard{\f21 Bosco, Luigi and Luigi Mittone (1997). \uc0\u8220 Tax Evasion and Moral Constraints: Some Experimental Evidence.\uc0\u8221 }{\f22\i Kyklos}{\f21 , vol 50, pp. 297-324.} \par\pard \par\pard{\f21 Brown-Kruse, Jamie and David Hummels (1993). \uc0\u8220 Gender Effects in Laboratory Public-Goods Contribution: Do Individuals Put Their Money Where Their Mouth Is? }{\f22\i Journal of Economic Behavior and Organization}{\f21 , vol 22, pp. 255-267.} \par\pard \par\pard{\f21 Buchan, Nancy, Eric Johnson and Rachel Croson (1998). \uc0\u8220 The Cultural Boundaries of Trust and Reciprocity in Economic Bargaining.\uc0\u8221 OPIM Working Paper, The Wharton School of the University of Pennsylvania. } \par\pard \par\pard{\f21 Burlando, R. and John Hey (1997). \uc0\u8220 Do Anglo-Saxons Free Ride More?\uc0\u8221 }{\f22\i Journal of Public Economics}{\f21 , vol 64, pp. 41-60.} \par\pard \par\pard{\f21 Camerer, Colin and Richard Thaler (1995). \uc0\u8220 Ultimatums, Dictators and Manners.\uc0\u8221 }{\f22\i Journal of Economic Perspectives}{\f21 , vol 9, pp. 209-219.} \par\pard \par\pard{\f21 Chakrabarti, Subir, William Lord and Peter Rangazas (1993). \uc0\u8220 Uncertain Altruism and Investment in Children.\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 83, pp. 994-1002.} \par\pard \par\pard{\f21 Clotfelter, C. (1985). }{\f22\i Federal Tax Policy and Charitable Giving}{\f21 . Chicago, IL: University of Chicago Press.} \par\pard \par\pard{\f21 Coate, Stephen (1995). \uc0\u8220 Altruism, the Samaritan\uc0\u8217 s Dilemma, and Government Transfer Policy.\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 85, pp. 46-57.} \par\pard \par\pard{\f21 Collard, David (1978). }{\f22\i Altruism and Economy}{\f21 . Oxford: Martin Robertson.} \par\pard \par\pard{\f21 Collard, David (1983). \uc0\u8220 Economics of Philanthropy: A Comment.\uc0\u8221 }{\f22\i Economic Journal}{\f21 , vol 93, pp. 637-638} \par\pard \par\pard\tx1260{\f23 Cooper, Russell, Douglas DeJong, Robert Forsythe and Thomas Ross (1996). \uc0\u8220 Cooperation without Reputation: Experimental Evidence from Prisoners\uc0\u8217 Dilemma Games.\uc0\u8221 }{\f24\i Games and Economic Behavior}{\f23 , vol 12, pp. 187-218.} \par\pard\tx1260 \par\pard\tx1260{\f23 Cornes, Richard and Todd Sandler (1984). \uc0\u8220 The Theory of Public Goods: Non-Nash Behaviour.\uc0\u8221 }{\f24\i Journal of Public Economics}{\f23 , vol 23, pp. 367-379.} \par\pard\tx1260 \par\pard\tx1260{\f23 Cox, Donald (1987). \uc0\u8220 Motives for Private Income Transfers.\uc0\u8221 }{\f24\i Journal of Political Economy}{\f23 , vol 95, pp. 508-546.} \par\pard\tx1260 \par\pard\tx1260{\f23 Cox, Donald and Mark Rank (1992). \uc0\u8220 Inter-vivos Transfers and Intergenerational Exchange.\uc0\u8221 }{\f24\i Review of Economics and Statistics, }{\f23 vol 74, pp. 305-314.} \par\pard\tx-720\tx0 \par\pard\tx-720\tx0{\f25 Croson, Rachel (1996). \uc0\u8220 Partners and Strangers Revisited.\uc0\u8221 }{\f26\i Economics Letters}{\f25 , vol 53, pp. 25-32.} \par\pard\tx-720\tx0 \par\pard\tx-720\tx0{\f25 Croson, Rachel (1997). \uc0\u8220 Effects of Eliciting Beliefs in a Linear Public Goods Game.\uc0\u8221 OPIM Working Paper, The Wharton School of the University of Pennsylvania.} \par\pard\tx-720\tx0 \par\pard\tx-720\tx0{\f25 Croson, Rachel (forthcoming). Feedback in Voluntary Contribution Mechanisms: An Experiment in Team Production. }{\f26\i Research in Experimental Economics}{\f25 .} \par\pard\tx1260 \par\pard\tx1260{\f23 Davis, Douglas and Charles Holt (1994). }{\f24\i Experimental Economics}{\f23 . Princeton, NJ: Princeton University Press.} \par\pard\tx1260 \par\pard\tx1260{\f23 Dudley, Dean (1993). \uc0\u8220 Individual Provision Choices in Voluntary Contribution Public Goods Environments: An Experimental Approach.\uc0\u8221 PhD thesis from Indiana University.} \par\pard\tx1260 \par\pard\tx1260{\f23 Eckel, Catherine and Philip Grossman (1996a). \uc0\u8220 The Relative Price of Fairness: Gender Differences in a Punishment Game.\uc0\u8221 }{\f24\i Journal of Economic Behavior and Organization,}{\f23 vol 30, pp. 143-158.} \par\pard\tx1260 \par\pard\tx1260{\f23 Eckel, Catherine and Philip Grossman (1996b). \uc0\u8220 Altruism in Anonymous Dictator Games.\uc0\u8221 }{\f24\i Games and Economic Behavior}{\f23 , vol 16, pp. 181-191.} \par\pard\tx1260 \par\pard\tx1260{\f23 --- (1996). }{\f24\i Forbes}{\f23 . Jan 22, v157, #2, p. 18. In Forbes Informer, edited by Kate Bohner Lewis.} \par\pard\tx1260 \par\pard\tx1260{\f23 Fehr, Ernst and Simon Gachter (forthcoming). \uc0\u8220 Reciprocity and Economics: The Economic Implications of }{\f24\i Homo Reciprocans}{\f23 .\uc0\u8221 Forthcoming, }{\f24\i European Economic Review.} \par\pard\tx1260 \par\pard\tx1260{\f23 Fehr, Ernst, Simon Gachter and Georg Kirchsteiger (1997). \uc0\u8220 Reciprocity as a Contract Enforcement Device: Experimental Evidence.\uc0\u8221 }{\f24\i Econometrica,}{\f23 vol 65, pp. 833-860.} \par\pard\tx1260 \par\pard\tx1260{\f23 Fehr, Ernst, Georg Kirchsteiger and Arno Riedl (1993). \uc0\u8220 Does Fairness Prevent Market Clearing?: An Experimental Investigation.\uc0\u8221 }{\f24\i Quarterly Journal of Economics}{\f23 , vol 108, pp. 437-459.} \par\pard\tx1260 \par\pard\tx1260{\f23 Fehr, Ernst, Georg Kirchsteiger and Arno Riedl (1996). \uc0\u8220 Involuntary Unemployment and Non-compensating Wage Differentials in an Experimental Labor Market.\uc0\u8221 }{\f24\i Economic Journal}{\f23 , vol 106, pp. 106-121.} \par\pard\tx1260 \par\pard\tx1260{\f23 Fehr, Ernst, Georg Kirchsteiger and Arno Riedl (1998). \uc0\u8220 Gift Exchange and Reciprocity in Competitive Experimental Markets.\uc0\u8221 }{\f24\i European Economic Review}{\f23 , vol 42, pp. 1-34.} \par\pard\tx1260 \par\pard\tx1260{\f23 Fehr, Ernst and Elena Tougareva (1996). \uc0\u8220 Do High Monetary Stakes Remove Reciprocal Fairness?: Experimental Evidence from Russia.\uc0\u8221 Working Paper, University of Zurich.} \par\pard\tx1260 \par\pard\tx1260{\f23 Fehr, Ernst and Jean-Robert Tyran (1996). \uc0\u8220 How do Institutions and Fairness Interact?\uc0\u8221 }{\f24\i Central European Journal for Operations Research and Economics}{\f23 , vol 4, pp. 69-84.} \par\pard\tx1260 \par\pard\tx1260{\f23 Feldstein, M. (1975). \uc0\u8220 The Income Tax and Charitable Contirbuitons: Part II--The Impact on Religious, Educational and Other Organizations.\uc0\u8221 }{\f24\i National Tax Journal}{\f23 , vol 28, pp. 81-100.} \par\pard\tx1260 \par\pard\tx1260{\f23 Fleishman, John (1988). \uc0\u8220 The Effects of Decision Framing and Others\uc0\u8217 Behavior on Cooperation in a Social Dilemma.\uc0\u8221 }{\f24\i Journal of Conflict Resolution}{\f23 , vol 32, pp. 162-180.} \par\pard\tx1260 \par\pard\tx1260{\f23 Forsythe, Robert, John Kennan and Barry Sopher (1991). \uc0\u8220 An Experimental Analysis of Strikes in Bargaining Games with One-Sided Private Information.\uc0\u8221 }{\f24\i American Economic Review}{\f23 , vol 81, pp. 253-278.} \par\pard\tx1260 \par\pard\tx1260{\f23 Forsythe, Robert, Joel Horowitz, Nathan Savin and Martin Sefton (1994). \uc0\u8220 Fairness in Simple Bargaining Experiments.\uc0\u8221 }{\f24\i Games and Economic Behavior,}{\f23 vol 6, pp. 347-369.} \par\pard\tx1260 \par\pard\tx1260{\f23 Frank, Richard and David Salkever (1991). \uc0\u8220 The Supply of Charity Services by Nonprofit Hospitals: Motives and Market Structure.\uc0\u8221 }{\f24\i Rand Journal of Economics}{\f23 , vol 22, pp. 430-445.} \par\pard\tx1260 \par\pard\tx1260{\f23 Frey, Bruno (1993). \uc0\u8220 Shirking or Work Morale? The Impact of Regulating.\uc0\u8221 }{\f24\i European Economic Review}{\f23 , vol 37, pp. 1523-1532.} \par\pard\tx1260 \par\pard\tx1260{\f23 Gachter, Simon and Armin Falk (1997). \uc0\u8220 Reputation or Reciprocity?\uc0\u8221 Working Paper, University of Zurich.} \par\pard\tx1260 \par\pard\tx1260{\f23 Gibbard, Allan and William Harper (1988). \uc0\u8220 Counterfactuals and Two Kinds of Expected Utility.\uc0\u8221 in }{\f24\i Decision, Probability and Utility: Selected Readings}{\f23 , Gardenfors and Sahlin (eds.), New York, NY: Cambridge University Press.} \par\pard\tx1260 \par\pard\tx1260{\f23 --- (1996). }{\f24\i Giving USA}{\f23 , Kaplan, Ann E. (ed.) New York, NY: American Association of Funding Counsel.} \par\pard\tx1260 \par\pard\tx1260{\f23 Greene, William (1990). }{\f24\i Econometric Analysis}{\f23 . New York, NY: Macmillan Publishing Company.} \par\pard\tx1260 \par\pard\tx1260{\f23 Guth, Werner, Peter Ockenfels and Markus Wendel (1997). \uc0\u8220 Cooperation Based on Trust: An Experimental Investigation.\uc0\u8221 }{\f24\i Journal of Economic Psychology}{\f23 , vol 18, pp. 15-43.} \par\pard\tx1260 \par\pard{\f21 Haltiwanger, John and Michael Waldman (1993). \uc0\u8220 The Role of Altruism in Economic Interaction.\uc0\u8221 }{\f22\i Journal of Economic Behavior and Organization}{\f21 , vol 21, pp. 1-15.} \par\pard \par\pard{\f21 Harsanyi, John (1980). \uc0\u8220 Rule Utilitarianism, Rights, Obligations and the Theory of Rational Behaviour.\uc0\u8221 }{\f22\i Theory and Decision}{\f21 , vol 12, pp. 115-133.} \par\pard \par\pard{\f21 Hayashi, Fumio (1995). \uc0\u8220 Is the Japanese Extended Family Altruistically Linked? A Test Based on Engel Curves.\uc0\u8221 }{\f22\i Journal of Political Economy,}{\f21 vol 103, pp. 661-674.} \par\pard \par\pard{\f21 Hoffman, Elizabeth, Kevin McCabe and Vernon Smith (1996). \uc0\u8220 Social Distance and Other-Regarding Behavior in Dictator Games.\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 86, pp. 653-660.} \par\pard \par\pard{\f21 Hoffman, Elizabeth, Kevin McCabe, Jason Shachat and Vernon Smith (1994). \uc0\u8220 Preferences, Property-Rights and Anonymity in Bargaining Games.\uc0\u8221 }{\f22\i Games and Economic Behavior}{\f21 , vol 7, pp. 346-380.} \par\pard \par\pard{\f21 Hollander, Heinz (1990). \uc0\u8220 A Social Exchange Approach to Voluntary Contribution.\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 80, pp. 1157-1167.} \par\pard \par\pard{\f21 Hood, R., S. Martin and Lars Osberg (1977). \uc0\u8220 Economic Determinants of Individual Charitable Donations in Canada.\uc0\u8221 }{\f22\i Canadian Journal of Economics}{\f21 , vol 10, pp. 653-669.} \par\pard\tx1260 \par\pard\tx1260{\f23 Hori, Hajime (1992). \uc0\u8220 Utility Functionals with Nonpaternalistic Intergenerational Altruism: The Case Where Altruism Extends to Many Generations.\uc0\u8221 }{\f24\i Journal of Economic Theory}{\f23 , vol 56, pp. 451-467.} \par\pard\tx1260 \par\pard{\f21 Isaac, R. Mark and James Walker (1988a). \uc0\u8220 Group Size Effects in Public Goods Provision: The Voluntary Contribution Mechanism.\uc0\u8221 }{\f22\i Quarterly Journal of Economics}{\f21 , vol 103, pp. 179-199.} \par\pard \par\pard{\f21 Isaac, R. Mark and James Walker (1988b). \uc0\u8220 Communication and Free-Riding Behavior: The Voluntary Contribution Mechanism.\uc0\u8221 }{\f22\i Economic Inquiry}{\f21 , vol 26, pp. 585-608.} \par\pard \par\pard{\f21 Isaac, R. Mark and James Walker (1991). \uc0\u8220 Costly Communication: An Experiment in a Nested Public Goods Problem.\uc0\u8221 in }{\f22\i Laboratory Research in Political Economy}{\f21 , Palfrey (ed.), Ann Arbor, MI: University of Michigan Press.} \par\pard \par\pard{\f21 Isaac, R. Mark and James Walker (1992). \uc0\u8220 Nash as an Organizing Principle in the Voluntary Contribution of Public Goods.\uc0\u8221 Working Paper, Indiana University. } \par\pard \par\pard{\f21 Isaac, R. Mark, James Walker and Arlington Williams (1994). \uc0\u8220 Group Size and the Voluntary Provision of Public Goods: Experimental Evidence Using Large Groups.\uc0\u8221 }{\f22\i Journal of Public Economics}{\f21 , vol 54, pp. 1-36.} \par\pard \par\pard{\f21 Jacobsen, Eva and Abdolkarim Sadrieh (1996). \uc0\u8220 Experimental Proof for the Motivational Importance of Reciprocity.\uc0\u8221 Discussion Paper B-386, University of Bonn.} \par\pard \par\pard{\f21 Jones, Philip (1996). \uc0\u8220 Rents from In-Kind Subsidy: \uc0\u8220 Charity\uc0\u8221 in the Public Sector.\uc0\u8221 }{\f22\i Public Choice}{\f21 , vol 86, pp. 359-378.} \par\pard \par\pard{\f21 Jones-Lee, Michael. (1992). \uc0\u8220 Paternalistic Altruism and the Value of Statistical Life.\uc0\u8221 }{\f22\i Economic Journal}{\f21 , vol 102, pp. 80-90.} \par\pard \par\pard\tx1260{\f23 Khanna, Jyoti, John Posnett and Todd Sandler (1995). \uc0\u8220 Charity Donations in the UK: New Evidence Based on Panel Data.\uc0\u8221 }{\f24\i Journal of Public Economics, }{\f23 vol 56, pp. 257-272.} \par\pard\tx1260 \par\pard\tx1260{\f23 Kirchler, Erich, Ernst Fehr and Robert Evans (1996). \uc0\u8220 Social Exchange in the Labor Market: Reciprocity and Trust versus Egoistic Money Maximization.\uc0\u8221 }{\f24\i Journal of Economic Psychology}{\f23 , vol 17, pp. 313-341.} \par\pard\tx1260 \par\pard\tx1260{\f23 Kogut, Bruce (1989). \uc0\u8220 The Stability of Joint Ventures: Reciprocity and Competitive Rivalry.\uc0\u8221 }{\f24\i Journal of Industrial Economics}{\f23 , vol 38, pp. 183-198.} \par\pard\tx1260 \par\pard\tx1260{\f23 Kranton, Rachel (1996). \uc0\u8220 Reciprocal Exchange: A Self-Sustaining System.\uc0\u8221 }{\f24\i American Economic Review}{\f23 , vol 86, pp. 830-851.} \par\pard\tx1260 \par\pard\tx1260{\f23 Kreps, David, Paul Milgrom, John Roberts and Robert Wilson (1982). \uc0\u8220 Rational Cooperation in the Finitely Repeated Prisoners\uc0\u8217 Dilemma.\uc0\u8221 }{\f24\i Journal of Economic Theory}{\f23 , vol 27, pp. 245-252.} \par\pard\tx1260 \par\pard{\f21 Laffont, Jean-Jacques (1975). \uc0\u8220 Macroeconomic Constraints, Economic Efficiency and Ethics: An Introduction to Kantian Economics.\uc0\u8221 }{\f22\i Economica}{\f21 , vol. 42, pp. 430-437.} \par\pard \par\pard{\f21 Laitner, John and F. Thomas Juster (1996). \uc0\u8220 New Evidence on Altruism: A Study of TIAA-CREF Retirees.\uc0\u8221 }{\f22\i American Economic Review}{\f21 , vol 86, pp. 893-908.} \par\pard \par\pard\tx1260{\f23 Ledyard, John (1995). \uc0\u8220 Public Goods: A Survey of Experimental Research.\uc0\u8221 in }{\f24\i The Handbook of Experimental Economics}{\f23 . Roth and Kagel, eds. Princeton, NJ: Princeton University Press.} \par\pard\tx1260 \par\pard{\f21 Lewis, David (1988). \uc0\u8220 Causal Decision Theory.\uc0\u8221 in }{\f22\i Decision, Probability and Utility: Selected Readings}{\f21 , Gardenfors and Sahlin (eds.), New York, NY: Cambridge University Press.} \par\pard \par\pard{\f21 Margolis, Howard (1982). }{\f22\i Selfishness, Altruism and Rationality}{\f21 . Cambridge: Cambridge University Press.} \par\pard\tx1260 \par\pard\tx1260{\f23 Marwell, Gerald and Ruth Ames (1979). \uc0\u8220 Experiments on the Provision of Public Goods I: Resources, Interest, Group Size and the Free-Rider Problem.\uc0\u8221 }{\f24\i American Journal of Sociology,}{\f23 vol 84, pp. 1335-1360..} \par\pard\tx1260 \par\pard\tx1260{\f23 Marwell, Gerald and Ruth Ames (1980). \uc0\u8220 Experiments on the Provision of Public Goods II: Provision Points, Stakes, Experience, and the Free Rider Problem.\uc0\u8221 }{\f24\i American Journal of Sociology,}{\f23 vol 85, pp. 926-937.} \par\pard\tx1260 \par\pard\tx1260{\f23 Marwell, Gerald and Ruth Ames (1981). \uc0\u8220 Economists Free Ride, Does Anyone Else?: Experiments on the Provision of Public Goods, IV.\uc0\u8221 }{\f24\i Journal of Public Economics,}{\f23 vol 15, pp. 295-310.} \par\pard\tx1260 \par\pard\tx1260{\f23 Messick, David, Henk Wilke, Marilynn Brewer, Roderick Kramer, Patricia English Zemke and Layton Lui (1983). \uc0\u8220 Individual Adaptations and Structural Change as Solutions to Social Dilemmas.\uc0\u8221 }{\f24\i Journal of Personality and Social Psychology,}{\f23 vol 44, pp. 294-309.} \par\pard\tx1260 \par\pard\tx1260{\f23 --- (1998). \uc0\u8220 Our Titanic Love Affair.\uc0\u8221 }{\f24\i Newsweek}{\f23 , February 23, p. 58 by David Anson.} \par\pard\tx1260 \par\pard{\f21 Nozick, Robert (1990). \uc0\u8220 Newcomb\uc0\u8217 s Problem and Two Principles of Choice.\uc0\u8221 in }{\f22\i Rationality in Action: Contemporary Approaches}{\f21 , Moser (ed.), New York, NY: Cambridge University Press.} \par\pard \par\pard{\f21 Offerman, Theo, Joep Sonnemans and Arthur Schram (1996). \uc0\u8220 Value Orientations, Expectations and Voluntary Contributions in Public Goods.\uc0\u8221 }{\f22\i Economic Journal}{\f21 , vol 106, pp. 817-845.} \par\pard \par\pard\tx1260{\f23 Palfrey, Thomas and Jeffrey Prisbrey (1996). \uc0\u8220 Altruism, Reputation and Noise in Linear Public Goods Experiments.\uc0\u8221 }{\f24\i Journal of Public Economics}{\f23 , vol 61, pp. 409-427.} \par\pard\tx1260 \par\pard\tx1260{\f23 Palfrey, Thomas and Howard Rosenthal (1988). \uc0\u8220 Private Incentives in Social Dilemmas: The Effects of Incomplete Information and Altruism.\uc0\u8221 }{\f24\i Journal of Public Economics}{\f23 , vol 35, pp. 309-332.} \par\pard\tx1260 \par\pard\tx1260{\f23 Poppe, Matthijs and Lisbeth Utens (1986). \uc0\u8220 Effects of Greed and Fear of being Gypped in a Social Dilemma Situation with Changing Pool Size.\uc0\u8221 }{\f24\i Journal of Economic Psychology,}{\f23 Vol 7, pp. 61-73.} \par\pard\tx1260 \par\pard{\f21 Posnett, John and Todd Sandler (1986). \uc0\u8220 Joint Supply and the Finance of Charitable Activity.\uc0\u8221 }{\f22\i Public Finance Quarterly,}{\f21 vol 14, pp. 209-222.} \par\pard \par\pard{\f21 Rangazas, Peter (1991). \uc0\u8220 Human Capital Investment in Wealth-Constrained Families with Two-Sided Altruism.\uc0\u8221 }{\f22\i Economics Letters, }{\f21 vol 35, pp. 137-141.} \par\pard \par\pard{\f21 Rapoport, Amnon and Dalit Eshed-Levy (1989). \uc0\u8220 Provision of step-level public goods: Effects of greed and fear of being gypped.\uc0\u8221 }{\f22\i Organizational Behavior and Human Decision Processes,}{\f21 vol 44, pp. 325-344.} \par\pard \par\pard{\f21 Rapoport, Amnon and Ramzi Suleiman (1993). \uc0\u8220 Incremental contribution in step-level public goods games with asymmetric players.\uc0\u8221 }{\f22\i Organizational Behavior and Human Decision Processes,}{\f21 vol 55, pp. 171-194.} \par\pard \par\pard{\f21 Roberts, Russell (1984). \uc0\u8220 A Positive Model of Private Charity and Public Transfers\uc0\u8221 }{\f22\i Journal of Political Economy,}{\f21 vol 92, pp. 136-148.} \par\pard\tx1260 \par\pard\tx1260{\f23 Roberts, Russell (1987). \uc0\u8220 Financing Public Goods.\uc0\u8221 }{\f24\i Journal of Political Economy}{\f23 , vol 95, pp. 420-437.} \par\pard\tx1260 \par\pard\tx1260{\f23 Rotemberg, Julio (1994). \uc0\u8220 Human Relations in the Workplace.\uc0\u8221 }{\f24\i Journal of Political Economy, }{\f23 vol 102, pp. 684-717.} \par\pard\tx1260 \par\pard\tx1260{\f23 Saijo, Toshi and H. Nakamura (1995). \uc0\u8220 The Spite Dilemma in Voluntary Contribution Mechanism Experiments.\uc0\u8221 }{\f24\i Journal of Conflict Resolution}{\f23 , vol 39, pp. 535-560.} \par\pard\tx1260 \par\pard\tx1260{\f23 Samuelson, Paul (1993). \uc0\u8220 Altruism as a Problem Involving Group versus Individual Selection in Economics and Biology.\uc0\u8221 }{\f24\i American Economic Review}{\f23 , vol 83, pp. 143-148.} \par\pard\tx1260 \par\pard\tx1260{\f23 Schroeder, David, Thomas Jensen, Andrew Reed, Debra Sullivan and Michael Schwab (1983). \uc0\u8220 The Actions of Others as Determinants of Behavior in Social Trap Situations.\uc0\u8221 }{\f24\i Journal of Experimental Social Psychology,}{\f23 vol 19, pp. 522-539.} \par\pard\tx1260 \par\pard{\f21 Schwartz, Robert (1970). \uc0\u8220 Personal Philanthropic Contributions.\uc0\u8221 }{\f22\i Journal of Political Economy}{\f21 , vol 78, pp. 1264-1291. } \par\pard \par\pard{\f21 Sefton, Martin and Richard Steinberg (1996). \uc0\u8220 Reward Structures in Public Goods Experiments.\uc0\u8221 }{\f22\i Journal of Public Economics}{\f21 , vol 61, pp. 263-287.} \par\pard \par\pard{\f21 Sell, Jane and Rick Wilson (1991). \uc0\u8220 Levels of Information and Contributions to Public Goods.\uc0\u8221 }{\f22\i Social Forces}{\f21 , vol 70, pp. 107-124.} \par\pard \par\pard{\f21 Sen, Amartya (1977). \uc0\u8220 Rational Fools: A Critique of the Behavioral Foundations of Economic Theory.\uc0\u8221 }{\f22\i Journal of Philosophy and Public Affairs, }{\f21 vol 6, pp. 317-344.} \par\pard \par\pard{\f21 Shafir, Eldar and Amos Tversky (1992). \uc0\u8220 Thinking Through Uncertainty: Nonconsequential Reasoning and Choice.\uc0\u8221 }{\f22\i Cognitive Psychology}{\f21 , vol 24, pp. 449-474.} \par\pard \par\pard{\f21 Smith, Vernon (1990). Experimental Economics: Behavioral Lessons for Microeconomic Theory and Applications.\uc0\u8221 Nancy L. Schwartz Memorial Lecture. Evanston, IL: Northwestern University, Kellogg Graduate School of Management.} \par\pard \par\pard{\f21 Smith, Vincent, Michael Kehoe and Mary Cremer (1995). \uc0\u8220 The Private Provision of Public Goods: Altruism and Voluntary Giving.\uc0\u8221 }{\f22\i Journal of Public Economics, }{\f21 vol 58, pp. 107-126.} \par\pard \par\pard{\f21 Solow, John (1994). \uc0\u8220 Paternalistic Preferences, Interpersonal Transfers and Reciprocity.\uc0\u8221 }{\f22\i Southern Economic Journal}{\f21 , vol. 61, pp. 379-386.} \par\pard \par\pard{\f21 Stigler, George (1974). \uc0\u8220 Free Riders and Collective Action: An Appendix to Theories of Economic Regulation.\uc0\u8221 }{\f22\i Bell Journal of Economics}{\f21 , vol 5, pp. 359-365.} \par\pard\tx1260 \par\pard\tx1260{\f23 Strawcyznski, Michel (1994). \uc0\u8220 Government Intervention as a Bequest Substitute.\uc0\u8221 }{\f24\i Journal of Public Economics}{\f23 , vol 53, pp. 477-495.} \par\pard\tx1260 \par\pard\tx1260{\f23 Struthers, John and Alistair Young (1989). \uc0\u8220 The Economics of Voting: Theories and Evidence.\uc0\u8221 }{\f24\i Journal of Economic Studies}{\f23 , vol 16, pp. 1-43.} \par\pard\tx1260 \par\pard\tx1260{\f23 Sugden, Robert (1984). \uc0\u8220 Reciprocity: The Supply of Public Goods through Voluntary Contributions.\uc0\u8221 }{\f24\i Economic Journal,}{\f23 vol 94, pp. 772-787.} \par\pard\tx1260 \par\pard\tx1260{\f23 Sugden, Robert (1985). \uc0\u8220 Consistent Conjectures and Voluntary Contributions to Public Goods: Why the Conventional Theory Does not Work.\uc0\u8221 }{\f24\i Journal of Public Economics}{\f23 , vol 27, pp. 117-124.} \par\pard\tx1260 \par\pard\tx1260{\f23 Suleiman, Ramzi and Amnon Rapoport (1992). \uc0\u8220 Provision of Step-level Public Goods with Continuous Contribution.\uc0\u8221 }{\f24\i Journal of Behavioral Decision Making}{\f23 , vol 5, pp. 133-153.} \par\pard\tx1260 \par\pard\tx1260{\f23 Tamura, Robert (1994). \uc0\u8220 Fertility, Human Capital and the Wealth of Families.\uc0\u8221 }{\f24\i Economic Theory}{\f23 , vol 4, pp. 593-603.} \par\pard\tx1260 \par\pard\tx1260{\f23 Tcha, MoonJoong (1995). \uc0\u8220 Altruism, Household Size and Migration.\uc0\u8221 }{\f24\i Economics Letters}{\f23 , vol 49, pp. 441-445.} \par\pard\tx1260 \par\pard\tx1260{\f23 Tversky, Amos and Eldar Shafir (1992). \uc0\u8220 The Disjunction Effect in Choice Under Uncertainty.\uc0\u8221 }{\f24\i Psychological Science,}{\f23 vol 3, pp. 305-309.} \par\pard\tx1260 \par\pard\tx1260{\f23 Unger, Lynette (1991). \uc0\u8220 Altruism as a Motivation to Volunteer.\uc0\u8221 }{\f24\i Journal of Economic Psychology}{\f23 , vol 12, pp. 71-100.} \par\pard\tx1260 \par\pard\tx1260{\f23 Van Huyck, John, Raymond Battalio and M Walters (1995). \uc0\u8220 Commitment versus Discretion in the Peasant-Dictator Game.\uc0\u8221 }{\f24\i Games and Economic Behavior}{\f23 , vol 10, pp. 143-170.} \par\pard\tx1260 \par\pard\tx1260{\f23 van Winden, Frans and Claudia Keser (1997). \uc0\u8220 Partners Contribute More to Public Goods than Strangers: Conditional Cooperation.\uc0\u8221 Working Paper, University of Amsterdam.} \par\pard\tx1260 \par\pard\tx1260{\f23 Vesterlund, Lise (1998). \uc0\u8220 The Informational Value of Sequential Fundraising.\uc0\u8221 Working Paper, Iowa State University.} \par\pard\tx1260 \par\pard\tx1260{\f23 Wade-Benzoni, Kimberly, Ann Tenbrunsel and Max Bazerman (1996). \uc0\u8220 Egocentric Interpretations of Fairness in Asymmetric, Environmental Social Dilemmas: Explaining Harvesting Behavior and the Role of Communication.\uc0\u8221 }{\f24\i Organizational Behavior and Human Decision Processes}{\f23 , vol 67, pp. 111-126.} \par\pard\tx1260 \par\pard\tx1260{\f23 Warr, Peter (1982). \uc0\u8220 Pareto Optimal Redistribution and Private Charity.\uc0\u8221 }{\f24\i Journal of Public Economics}{\f23 , vol 18, pp. 131-138.} \par\pard\tx1260 \par\pard\tx1260{\f23 Warr, Peter (1983). \uc0\u8220 The Private Provision of a Public Good is Independent of the Distribution of Income.\uc0\u8221 }{\f24\i Economics Letters}{\f23 , vol 13, pp. 207-211.} \par\pard\tx1260 \par\pard\tx1260{\f23 Weimann, Joachim (1994). \uc0\u8220 Individual Behaviour in a Free Riding Experiment\uc0\u8221 }{\f24\i Journal of Public Economics,}{\f23 vol 54, pp. 185-200.} \par\pard\tx1260 \par\pard\tx1260{\f23 Wilhelm, Mark (1996). \uc0\u8220 Bequest Behavior and the Effect of Heirs Earnings: Testing the Altruistic Model of Bequests.\uc0\u8221 }{\f24\i American Economic Review}{\f23 , vol 86, pp. 874-892.} \par\pard\tx1260 \par\pard\tx1260{\f23 Wilson, Rick and Jane Sell (1997). \uc0\u8220 Liar, Liar: Cheap Talk and Reputation in Repeated Public Goods Settings.\uc0\u8221 }{\f24\i Journal of Conflict Resolution, }{\f23 vol 41, pp. 695-717.} \par\pard\tx1260{\f23\page} \par\pard\tx1260{\f23\page} \par\pard\tx1260{\f23\page} \par\pard\tx1260{\f23\page} \par\pard\tx1260 \par\pard\ri180\sl480\slmult1 \par\pard\ri180\sl480\slmult1{\f8\page} \par\pard\ri180\sl480\slmult1{\f8\page} \par\pard\ri180\sl480\slmult1{\f8\page} \par\pard\ri180\sl480\slmult1{\f8\page} \par\pard\ri180\sl480\slmult1{\f8\page} \par\pard\ri180\sl480\slmult1{\f8\page} \par\pard\ri180\sl480\slmult1{\f8\page} \par\pard\ri180\sl480\slmult1{\f8\page} \par\pard\ri180\sl480\slmult1{\f8\page\page}{\f7\b Footnotes} \par\pard\ri180\sl480\slmult1 \par\pard \par\pard \par\pard \par\pard \par\pard\ri360\tqc\tx4320\tqc\tx8640{\f27 30} \par\pard\qc\ri360\tqc\tx4320\tqc\tx8640{\f28 30} \par\pard \par\pard\qc\ri360\tqc\tx4320\tqc\tx8640 \par\pard \par\pard \par\pard\tx720\tx1440\tx7200{\f29 The actual amount contributed by individuals in 1995 was $116,230,000,000. This number excludes charitable giving by corporations, foundations and bequests.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 That is, multiple agents can consume the good at the same time (nonrival) and it is not possible to exclude agents who did not pay for the good from consuming it (nonexcludable). } \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 The multiple P is often called the marginal per capita return (MPCR) and is the marginal return to each individual on a contribution of one unit to the group account (Isaac and Walker, 1988a).} \par\pard\tx720\tx1440\tx7200 \par\pard\ri360\tx720\tx1440\tx7200{\f30 Other studies using this mechanism have examined the effect of relative payoffs (Isaac and Walker, 1988a) communication (Isaac and Walker, 1988b, 1991; Wilson and Sell, 1997), culture (Burlando and Hey, 1997), reputation and learning (Andreoni, 1988b; Weimann, 1994; Croson, 1996; van Winden and Keser, 1997), random noise (Andreoni, 1995; Palfrey and Prisbrey, 1996), payments (Sefton and Steinberg, 1996), spite (Saijo and Nakamura, 1995), group size (Isaac, Walker and Williams, 1994), discreteness of the public good (Asch, Gigliotti and Polito, 1993), and information (Sell and Wilson, 1991; Croson, forthcoming), as well as many other issues.} \par\pard\ri360\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 A similar technique of belief elicitation has been used in public goods games in a slightly different context. In these studies, voluntary contribution mechanisms are run which have interior Nash equilibria (rather than a boundary equilibrium). Authors then elicit subjects\uc0\u8217 beliefs about others\uc0\u8217 behavior and categorizes subjects based on whether they play best responses to their own beliefs (see e.g. Isaac and Walker, 1992; Dudley 1993; Offerman, Sonnemans and Schram, 1996; Suleiman and Rapoport, 1992; Rapoport and Eshed-Levy, 1989; Rapoport and Suleiman, 1993; ).} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 More recent work in philosophy (e.g. Gibbard and Harper, 1988; Lewis, 1988; Nozick, 1990) and psychology (e.g. nonconsequential reasoning of Shafir and Tversky, 1993 and Tversky and Shafir, 1992) suggests that these sorts of misperceptions of causality both can be and are used to solve many social and individual decision problems.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 Laffont (1975) also discusses the social benefits of a government convincing the population that this belief is true.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 In fact, in this linear case, x}{\f31\sub i}{\f29 * = E}{\f31\sub i}{\f29 ,} \par\pard\tx720\tx1440\tx7200 \par\pard\ri360\tx720\tx1440\tx7200{\f30 Both Collard (1978) and Becker (1974) show that altruism need not lead to an infinite explosion of utility between multiple altruistic individuals as long as an individual\uc0\u8217 s own utility (or consumption) is more important to him than anothers\uc0\u8217 .} \par\pard\ri360\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 That is, each dollar increase in government grants should result in a dollar decrease in private giving.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 A number of interesting experimental studies have investigated gender differences in altruistic preferences with different results. Bolton and Katok (1995) showed no significant differences while other studies (Andreoni and Vesterlund, 1997; Eckel and Grossman, 1996a; Brown-Kruse and Hummels, 1993) demonstrate significant differences.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 Other theories of altruism also predict this negative correlation (e.g. Arrow, 1975; Sen, 1977; Roberts, 1984, 1987; Posnett and Sandler, 1986; Margolis, 1982; Schwartz, 1970; Hood, Martin and Osberg, 1977; Collard, 1978).} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 It is worth noting that such a positive correlation has been assumed in theories of voluntary activities (e.g. Cornes and Sandler, 1984).} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 Because they estimated the contributions of the other three members of the group, subjects could not influence the accuracy of their guess by strategically changing their own contribution.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 In addition to their earnings from the public good, subjects earned 50\'a2 if their estimate was exactly right. If their estimate was a bit off, they earned 25\'a2 divided by the (absolute) distance between their estimate and the true contribution. This payment scheme leads to an approximation of a single-peaked curve.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 Croson (1997) compares contributing behavior between this treatment and a traditional linear public goods experiment treatment.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 In addition there were three practice periods before the first game began to familiarize subjects with the computer program and the process. Subjects were not paid their earnings during the practice periods and no practice periods were run before the second game. Raw data as well as the instructions used are available from the author. } \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 The same regression including dummy variables for each group had similar results, as did one including dummies for the period numbers and a two-factor random effects regression (individual and period).} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 Notice we have two observations for each individual, one for period 1 of the original game and one for period 1 of the restart game. Thus we include the individual dummies and random effects.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 Identical regressions without the individual dummy variables yield similar results (the period 1 coefficient on GUESS is .192, p < .01, the period 10 coefficient on GUESS is .184, p < .01). Identical regressions run for each of the games separately on period 1 play also yield similar results (for the original game, the period 1 coefficient on GUESS is .215, p < .01 and for the restart game, the same coefficient is .313, p < .01).} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 A similar test excluding the observations of zero correlation was also run. The null hypothesis of equal number of positive and negative correlations was rejected at p<.01.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 The hypothesis that the mean of the distribution of errors in the first game is equal to zero can be rejected using a t-test at the 5% level for eight out of 24 subjects. It cannot be rejected for any subjects in the second game.} \par\pard\tx720\tx1440\tx7200 \par\pard\tx720\tx1440\tx7200{\f29 A similar test excluding the observations of zero contributions was also run. The null hypothesis of equal number of positive and negative correlations was rejected at p<.01 for all four treatments as well as the combined data.} \par\pard\tx720\tx1440\tx7200 \par\pard\ri360\tx720\tx1440\tx7200{\f30 Remember that Sugden\uc0\u8217 s notion of reciprocity is a simultaneous rather than a sequential one. It\uc0\u8217 s not that subjects in this experiment are \uc0\u8220 rewarding\uc0\u8221 their group members for past performance. Instead, they are trying to \uc0\u8220 match\uc0\u8221 the contributions they expect of others in their group.} \par\pard\ri360\tx720\tx1440\tx7200 \par\pard\ri360\tx720\tx1440\tx7200{\f30 Notice it is quite possible for an individual\uc0\u8217 s contribution to be above the maximum (or below the minimum) of the other three members of his group. In fact, the contributions one person in each group in each period will have this characteristic.} \par\pard\ri360\tx720\tx1440\tx7200 \par\pard\ri360\tx720\tx1440\tx7200{\f30 A more recent example of reciprocal behavior is found in the entertainment industry. During the making of }{\f32\i Titanic}{\f30 , when the film was running over budget and failed to make the planned release date, director James Cameron voluntarily gave up fees for the film as well as his percentage of the profits in order to assuage concerns of executives at Twentieth Century Fox and Paramount Picture Studios. Now that }{\f32\i Titanic}{\f30 is the third-biggest grossing film of all time, those waived fees and percentages total to approximately $50 million. A report in }{\f32\i Newsweek }{\f30 (1998) says that studio heads are now considering voluntarily and unilaterally paying Cameron the very fees he had agreed to waive.} \par\pard \par\pard\ri180\sl480\slmult1}