Tax Morale and Social Desirability Bias: Examining the Shadow Economy Experimentally

Informations projet

TAXMORALBIAS

N° de convention de subvention: 101109679

DOI

10.3030/101109679

Projet clôturé

Date de signature de la CE 8 Mai 2023

Date de début 16 Octobre 2023

Date de fin 15 Octobre 2025

Financé au titre de

• Marie Skłodowska-Curie Actions (MSCA)

Coût total

Aucune donnée

Contribution de l’UE

€ 188 590,08

Coordonné par

FONDAZIONE COLLEGIO CARLO ALBERTO
Italy

Periodic Reporting for period 1 - TAXMORALBIAS (Tax Morale and Social Desirability Bias: Examining the Shadow Economy Experimentally)

Période du rapport: 2023-10-16 au 2025-10-15

Résumé du contexte et des objectifs généraux du projet

The TAXMORALBIAS project investigated the relationship between tax morale and social desirability bias to advance understanding of the shadow economy and tax compliance in Europe using experimental methods. Tax morale — the shared belief in the inherent good of contributing to society through tax payments — is widely recognized as a key factor explaining tax compliance. However, the project posed a novel research question: could tax morale itself generate social desirability bias, leading individuals to misrepresent their tax behaviour in surveys and resulting in biased estimates of shadow economy participation? This question is particularly relevant in Europe, where tax morale varies significantly across regions. Nordic countries display high tax morale coupled with reported high tax compliance, Southern Europe exhibits high tax morale alongside lower compliance, and Baltic countries show both lower tax morale and lower compliance levels. Understanding these dynamics is crucial for developing effective policies to combat the shadow economy and improve tax compliance. The project addressed this research question through a methodologically novel two-stage experimental design combining list experiments with vignette experiments conducted across three European countries representing North, East, and South Europe. In addition, the project included a large-scale field experiment conducted in collaboration with Latvia's State Revenue Service. This field experiment tested the integration of gamification elements in tax administration feedback to improve timely tax compliance among legal entities.

The project's conclusions underscore the value of experimental methods for understanding tax compliance behaviour and designing effective policy interventions. The evidence generated contributes to advancing shadow economy studies both methodologically and conceptually, offering new inquiry approaches grounded in behavioural science. The findings support policies that leverage behavioural insights, including gamification and personalized messaging, to improve tax administration effectiveness and citizen engagement. The research demonstrates that tax morale's relationship with tax behaviour is more complex than traditionally understood, influenced not only by individual moral convictions but also by how these convictions shape information disclosure in surveys and responses to institutional communications. By bridging laboratory research with field experiments and combining quantitative with qualitative insights, the project provides evidence-based foundations for future policy development in tax compliance and shadow economy reduction across Europe.

Travail effectué depuis le début du projet jusqu’à la fin de la période considérée dans le rapport et principaux résultats atteints jusqu’à présent

he project executed a comprehensive investigation of tax morale, social desirability bias, and tax compliance behaviour. The first component involved a cross-national list experiment combined with vignette priming, conducted across three European countries (representing North, East, and South regions) to examine how tax morale influences the disclosure of truthful information about tax evasion. List experiments, traditionally used in crime studies, provide a method to estimate sensitive behaviours like tax evasion by comparing response patterns between control and treatment groups, thereby identifying the degree of social desirability bias. The project enhanced this methodology by integrating vignette experiments—randomizing respondents to receive priming treatments about tax morale before completing the list experiment—enabling causal identification of tax morale's effect on the disclosure of truthful responses regarding tax evasion. This combination was novel within European tax compliance research and offered significant potential to deliver deeper understanding of both social desirability bias and tax morale's true role in shaping tax behaviour. This novel integration enabled identification of interaction effects between priming and individual tax morale levels — a methodologically innovative contribution to tax compliance research.

The second major component was a large-scale field experiment implemented in collaboration with Latvia's State Revenue Service, investigating the effectiveness of gamification in improving corporate tax compliance. The field experiment employed a randomization procedure to assign approximately five thousand taxpayers with histories of late employer report submissions to one of five experimental groups: a control group and four treatment groups receiving distinct gamified messages. Treatment arms incorporated theoretically grounded behavioural interventions including simple reminder messages, goal-framing communications emphasizing achievement of an A-rating status, achievement-based messaging with virtual tokens awarded for consecutive timely submissions, and collective commitment appeals referencing national budget progress.

The list experiment revealed that social desirability bias in tax evasion surveys manifests in a context-dependent manner rather than following universal patterns. Contrary to expectations of systematic underreporting, respondents demonstrated strategic calibration of disclosure across different question formats, with significant heterogeneity across countries. The integration of vignette-based norm priming did not substantially alter aggregate prevalence estimates, yet it did reduce inconsistent response patterns, suggesting that normative information influences how individuals strategically align their answers across survey methodologies. The placebo list experiment validated the methodological approach overall, confirming that observed response irregularities were specific to the sensitive tax domain rather than reflecting general design failures.

The field experiment on gamification demonstrated that behavioural interventions can effectively improve compliance with administrative requirements. Simple repeated communication significantly enhanced timely submission rates across all treatment conditions. Notably, goal-framing messages offering concrete, observable recognition proved more effective than intrinsic motivational appeals based on achievement or abstract collective commitments. These findings indicate that individual, tangible incentives coupled with social recognition mechanisms drive behavioural change more substantially than complex gamification tools lacking material or social salience, and that intervention effectiveness depends critically on alignment between the incentive structure and respondents' existing motivational landscape.

The project generated multiple scientific outputs documenting both methodological innovation and empirical findings. Manuscript examining strategic misreporting and social desirability bias through integrated list-vignette experiments has been submitted to the Journal of Economic Psychology, a second on gamification integration in tax compliance is being prepared for submission to the Journal of Public Economics. Beyond peer-reviewed publications, the project prepared a policy brief synthesizing findings on behavioural insights for tax administration through the Forum for Research on Eastern Europe and Emerging Economies (FROGEE).

Progrès au-delà de l’état des connaissances et impact potentiel prévu (y compris l’impact socio-économique et les conséquences sociétales plus larges du projet jusqu’à présent)

The project advanced the methodological frontier in measuring sensitive behaviours through a novel integration of list experiments with vignette-based norm priming. Prior research had employed list experiments independently or vignettes separately, but their combination enabled direct causal identification of how social norms influence response bias in surveys about illegal or stigmatized behaviours. This methodological innovation revealed that social desirability bias operates through multiple, interacting mechanisms — including strategic respondent behaviour, normative calibration, and context-specific threshold effects — rather than through a single, uniform underreporting process. The findings challenge conventional assumptions embedded in indirect questioning techniques and demonstrate that survey design itself constitutes a form of informational treatment capable of reshaping disclosure patterns. Beyond tax evasion specifically, this methodological approach is applicable to any domain involving sensitive self-reported behaviours where normative contexts vary substantially across populations, including illicit substance use, corruption, undeclared labour, and stigmatized health or financial practices.

The field experiment on gamification in tax administration extends beyond existing laboratory evidence by demonstrating that behavioural interventions operate under fundamentally different constraints in real-world administrative contexts compared to controlled experimental settings. Specifically, the finding that goal-framing with tangible, publicly observable recognition outperforms intrinsic motivation mechanisms contradicts prominent laboratory research suggesting that achievement-based gamification transcends material incentives. This real-world validation reveals that institutional design cannot simply transfer laboratory-validated interventions without attending to the underlying incentive ecology and social recognition mechanisms operative in specific administrative contexts. The results provide concrete evidence that public administration can leverage behavioural science to improve compliance with minimal cost, using simple communication modifications rather than expensive institutional restructuring. These findings establish a template for government-academic collaboration in conducting rigorous, large-scale experiments within operational administrative systems, demonstrating feasibility of randomized controlled trials at scale in public sector contexts.

Key needs for further research and uptake include systematic investigation of the generalizability of both findings across different administrative and cultural contexts. For the survey methodology, replication across additional countries and sensitive domains would establish whether the observed patterns of strategic misreporting and norm-priming effects represent universal survey dynamics or emerge specifically under particular socioeconomic and normative conditions. For the gamification intervention, testing variations in intervention design — such as alternative framing mechanisms, timing of message delivery, and integration with digital platforms — would clarify which design elements prove most cost-effective across diverse populations and compliance challenges. Development of practical guidance for tax administrations and other public agencies on implementing behavioural interventions would require demonstration projects in additional jurisdictions, supported by capacity-building workshops for policymakers and practitioners. International coordination and standardization of approaches to behavioural public administration, facilitated through forums such as the OECD's behavioural insights network, would accelerate uptake and enable comparative learning.

List experiment results on tax evasion: indirect vs. direct question