Insights from experiments with duopoly games: rational incremental decision-making

ABSTRACT

This study explores rational incremental decision-making, one of the most powerful principles in government decision-making. Although ‘muddling through’ is applicable in many decision-making situations, it has been argued that incremental decisions are rational in some conditions. This study replicates the game experiment using two-person games with the factorial design of uncertainty and risk in Hong Kong and Korea. Findings show that incremental decisions in prior rounds lead to rational results and thereby better rewards, and that uncertainty and risk slow down the process of reaching the Nash equilibrium. The conclusion discusses implications of rational incrementalism.

KEYWORDS:

• Signalling game
• rational decision-making
• incrementalism
• experiment
• replication

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. Tsang and Kwan (1999) outlined a typology of replication types based upon two criteria: different or same measurement and/or analysis and same data set, same population or different population. This typology delivers six replication types: (1) checking of analysis using the same data set and measures, (2) reanalysis of data, where new analytical tools are applied to an existing data set, (3) theory testing, where the same research design, measurement and analysis is implemented in the same population but using a different sample, (4) conceptual extension, whereby procedures vary but the same population is sampled, (5) empirical generalization, which uses the same design, measures and analysis, but in a different population and (6) generalizations and extensions, which involve new measurement, analysis and populations.

2. In the experiment, a player was provided with a plain game sheet without any markings or font changes.

3. Although the incremental strategy was included in the original hypotheses separately, this incremental hypothesis simply described decisions made more incrementally than in the other strategies due to the difficulty (high risk or high uncertainty) of selection.

4. (12, 14) and (14, 12) are (l, n) and (n, l) in the replication settings.

5. One player made a sequence of ‘l’–‘n’–‘l’–‘n’ selections, but the second player kept selecting ‘n’, so the player lost HK$2 in a few rounds and could not take bigger rewards than the equilibrium.

6. Based on the original study, we offered the following instruction: ‘Your game reward sheet is below. You pick rows. Depending on what your partner plays, you receive the value in the box. For instance, if you picked “a” and your partner also picked “a”, you would receive $5. After each play, you will get to see your partner’s play history’.

7. The Korean rewards were transformed into Hong Kong dollars for analysis.

8. Regression analysis with both the Hong Kong and Korean data also shows the significant effect of uncertainty, convergence and an incremental start in the first five rounds.

Additional information

Funding

This work was supported by the National Research Foundation of Korea: [NRF-2014S1A3A2044630].

Notes on contributors

M. Jin Lee

M. Jin Lee is research fellow of Public Policy at City University of Hong Kong. She is a core member of Laboratory for Public Management and Policy, and manages the experimental research part. Her research focuses on public management, decision-making, motivation and social network analysis.

M. Jae Moon

M. Jae Moon is Underwood Distinguished Professor of Public Administration at Yonsei Unive rsity. He is an elected fellow of National Academy of Public Administration and International Director of American Society for Public Administration. He also chairs the Research Committee of Korean Association of Public Administration. His research interests include e-government, public management and comparative administration.

Jungsook Kim

Jungsook Kim is PhD student of Public Administration at Yonsei University. Her research interests include non-profit organizations, governance and local finance.