The use of agent-based modelling to investigate tax compliance


Agent-based modelling can be used to investigate the behavioural and social aspects of tax compliance. We illustrate the approach with two models. The first model emphasises the role of occupational choice in tax compliance, and explores the effect of non-compliance on risk-taking and income distribution. The modelling of the compliance decision is discussed with an emphasis on decision-making under uncertainty and social interaction. We then add to the model a social network which governs the transmission of information on attitudes and beliefs, and investigate alternative audit strategies. A strategy of auditing a fixed number of taxpayers from each occupation dominates alternative strategies (including random and focussed strategies) in the sense of first-order stochastic dominance.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10


  1. 1.

    The computer code for the simulations we report is available on the SpringerPlus website. The code is written for Matlab but can be easily converted to run with Scilab.

  2. 2.

    The advantage of the beta distribution is in the flexible choice of parameters, allowing density functions with required skewness to be obtained, and the finite support ensures robust convergence of the numerical integration.

  3. 3.

    A matrix that is not symmetric captures uni-directional links. This can be used to investigate the effect of a “celebrity”.

  4. 4.

    It is assumed tax authority knows that in paid employment income tax is fully deducted at source, and there is no opportunity for earning additional income that could be concealed. This assumption could be modified in a more general model to allow an additional income for individuals in employment and a possibility to evade tax on that income.

  5. 5.

    In this context, strategy \(A\) dominates strategy \(B\) in the sense of first-order stochastic dominance, if for every level of revenue, \(R\), the probability of collecting at least \(R\) is higher under \(A\) than under \(B\). Equivalently, the empirical cdf of revenues collected under \(A\) is everywhere below (or to the right from) the empirical cdf of revenues collected under \(B\).


  1. Allingham M, Sandmo A (1972) Income tax evasion: a theoretical analysis. J Public Econ 1:323–338

    Article  Google Scholar 

  2. Al Nowaihi A, Dhami S (2007) Why do people pay taxes: expected utility theory versus prospect theory. J Econ Behav Organ 64:171–192

  3. Andrei AL, Comer K, Koelher M (2014) An agent-based model of network effects on tax compliance and evasion. J Econ Psychol 40:119–133

    Article  Google Scholar 

  4. Bernasconi M, Zanardi A (2004) Tax evasion, tax rates and reference dependence. FinanzArchiv 60:422–445

    Article  Google Scholar 

  5. Black J, de Meza D (1997) Everyone may benefit from subsidising entry to risky occupations. J Public Econ 66:409–424

    Article  Google Scholar 

  6. Bloomquist KM (2004) Multi-agent based simulations of the deterrent effects of taxpayer audits. In: Kalambokidis L (ed) Proceedings of the 97th annual conference on taxation. National Tax Association, Washington, DC, pp 159–173

    Google Scholar 

  7. Bloomquist KM (2012) Incorporating indirect effects in audit case selection: an agent-based approach. In: Plumley A (ed) 2012 IRS research bulletin. IRS, Washington, DC, pp 103–116

    Google Scholar 

  8. Chateauneuf A (1994) Modeling attitudes towards uncertainty and risk through the use of Choquet integral. Ann Oper Res 52:3–20

    Article  Google Scholar 

  9. Davis JS, Hecht G, Perkins JD (2003) Social behaviors, enforcement, and tax compliance dynamics. Account Rev 78:39–69

    Article  Google Scholar 

  10. Guala F, Mittone L (2005) Experiments in economics: external validity and the robustness of phenomena. J Econ Method 12:495–515

    Article  Google Scholar 

  11. Hashimzade N, Myles GD, Page F, Rablen MD (2014) Social networks and occupational choice: the endogenous formation of attitudes and beliefs about tax compliance. J Econ Psychol 40:134–146

    Article  Google Scholar 

  12. Hashimzade N, Myles GD, Tran-Nam B (2013) Applications of behavioural economics to tax evasion. J Econ Surv 27:941–977

    Google Scholar 

  13. Kahneman D, Tversky A (1979) Prospect theory: an analysis of decision under risk. Econometrica 47:263–293

    Article  Google Scholar 

  14. Kanbur SM (1981) Risk taking and taxation: an alternative perspective. J Public Econ 15:163–184

    Article  Google Scholar 

  15. Korobow A, Johnson C, Axtell R (2007) An agent-based model of tax compliance with social networks. Natl Tax J 60:589–610

    Google Scholar 

  16. Myles GD, Naylor RA (1996) A model of tax evasion with group conformity and social customs. Eur J Polit Econ 12:49–66

    Article  Google Scholar 

  17. Pestieau P, Possen UM (1991) Tax evasion and occupational choice. J Public Econ 45:107–125

    Article  Google Scholar 

  18. Piolatto A, Rablen MD (2013) Prospect theory and tax evasion: a reconsideration of the Yitzhaki Puzzle. IZA Discussion Paper No. 7760, IZA, Bonn

  19. Quiggin J (1982) A theory of anticipated utility. J Econ Behav Organ 3:323–343

    Article  Google Scholar 

  20. Slemrod J (2007) Cheating ourselves: the economics of tax evasion. J Econ Perspect 21:25–48

    Article  Google Scholar 

  21. Snow A, Warren RS Jr (2005) Ambiguity about audit probability, tax compliance, and taxpayer welfare. Econ Inq 43:865–871

    Article  Google Scholar 

  22. Tesfatsion L (2006) Agent-based computational economics: A constructive approach to economic theory. In: Tesfatsion L, Judd KL (eds) Handbook of computational economics, vol II. North-Holland, Amsterdam, pp 831–880

    Google Scholar 

  23. Yaniv G (1999) Tax compliance and advance tax payments: a prospect theory analysis. Natl Tax J 52:753–764

    Google Scholar 

  24. Yitzhaki S (1974) A note on income tax evasion: a theoretical analysis. J Public Econ 3:201–202

    Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Gareth D. Myles.

Additional information

Thanks are due to the ESRC for financial support under grant ES/K005944/1. Previous versions of the paper were presented at the Ottawa Workshop on Compliance and at shadow2013 in Münster. Gareth Myles worked on the paper during a visit to Bogazici University; their hospitality is appreciated.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Hashimzade, N., Myles, G.D., Page, F. et al. The use of agent-based modelling to investigate tax compliance. Econ Gov 16, 143–164 (2015).

Download citation


  • Agent-based modelling
  • Tax evasion
  • Attitudes
  • Beliefs
  • Social network
  • Occupational choice

JEL Classification

  • H26
  • D85
  • C63