Follow the First Follower: Donations to Charity and Pro-Social Behaviour

Leadership is present in wide variety of human interactions from organisational hierarchy to online charitable fundraising. Consequently, the subject has been studied in a multitude of fields from biology to history. In behavioural economics, the existence of leadership has been tested experimentally as a potential solution to social dilemmas with varying success. However, the majority of previous experimental studies have examined only one form of leadership where leadership where only one group member can exhibit it. In psychological research, there is an evolving interest in the concept of shared leadership, where leadership is shared among multiple group members. Likewise, Derek Sivers (2010) emphasised the importance of a first follower, who is a person who acts directly after the leader and makes the leader's actions more credible. This thesis tests the impact of first followership on solving social dilemmas and examines further the topic of shared leadership.

Chapter 2 studies whether the introduction of a first follower increases total investment in a linear public good game both theoretically and experimentally. Four investment sequences are considered including two new additions to the literature: the First Follower game and the Two Leader game. The already established Leader game and Sequential game are implemented as control treatments. The First Follower game has three investment stages; in the first stage, a leader invests, in the second stage a first follower invests and the rest of the group simultaneously invest in the third stage. The Two Leader game has only two investment stages; in the first stage, two leaders simultaneously invest in the first stage and then the rest of the group simultaneously invest in the second stage. The four investment sequences are ranked according to their expected total investment. No one investment sequence is always found to be superior, and thus the ranking is dependent on the strategic behaviour of early movers, namely leaders and first followers. In the experiment, the First Follower game had the highest total invested in the public good and Two Leader, the second highest. These results suggest that shared leadership or first followership may be superior to a one leader structure. Specifically, in the First Follower game, leaders set a better example by investing significantly more than leaders in other treatments, and first followers exhibit higher levels of reciprocity than later followers. All investment sequences were imposed exogenously to create a simple initial testing ground for introducing the first follower.

Chapter 3 expands on the endogenous leadership literature by comparing the efficiency of a game with two investment stages, a two-day treatment to a game with three investment stages, a three-day treatment. Using a novel experimental design, each group member chooses what investment stage they would like to invest during. Therefore, the number of group members investing on day one, equivalent to leaders, is unrestricted, hence one can also explore how many group members are willing to lead. Willingness to lead is greater in both treatments than observed in previous studies. Nevertheless, the proportion of leaders is significantly higher in the three-day treatment, this result leads to significantly higher total investment in the three-day treatment. Leader investment and total investment in these treatments, where there is endogeneity over the timing of investment choices is compared to the treatments from Chapter 2 where timing is exogenously imposed. Endogenous leaders are found not to invest significantly more than exogenous leaders, and similarly there is no significant difference in total investment. Nevertheless, the treatments with three investment stages, namely three-day and First Follower, have the highest levels of leader investment and overall total investment.

Chapter 4 extends the design of Corazzini, Cotton and Valbonesi (2015) to consider a real-time environment; this adjustment makes the game more akin to real world crowdfunding and competitive lobbying (Austen-Smith and Wright, 1992). Like Chapter 3, individuals choose when they would like to invest rather than a set exogenous investment sequence imposed on them as in Chapter 2. Two treatments are executed, a multiple public good treatment with four identical threshold public goods and another with only one threshold public good. It is hypothesised that the real-time environment increase efficiency in reaching a threshold in the presence of multiple threshold public goods. The experimental evidence implies otherwise; the multiple public good treatment leads to significantly lower levels of efficiency stemming from the inefficiency caused by group members investing in multiple projects when only one project can be fully funded.