Understanding drivers when investing for impact: an experimental study

Impact investing has expanded rapidly, from an estimated $4.3B in 2011 to $502B in 2018, with potential to reach $1T. The Global Impact Investing Network defines impact investments as those made with the intention to generate social and environmental impact alongside financial returns. While research has focused on impact assessment and measurement, less attention has been given to who chooses impact funds and why. This experiment-based study addresses this gap by examining behavioral drivers and socio-demographic correlates of choosing impact investing funds (IIF) over traditional investing funds (TIF). The study’s research questions are: (RQ1) Does previous knowledge about impact investing affect investment choices? (RQ2) How do preferences change with different framings and presentations of investment instruments? (RQ3) How much financial return are investors willing to sacrifice for social impact under different risk conditions? (RQ4) Do external incentives (e.g., fiscal benefits) affect choices? Findings aim to inform behavioral economics and policy/industry strategies to mainstream impact investing.

The paper situates impact investing within a growing scholarly and practitioner discourse emphasizing measurement of social and environmental outcomes. Prior work has developed frameworks for impact assessment and accountability, while CSR and socially responsible consumption literatures explore reputation- and brand-related behaviors. However, socio-demographic and behavioral determinants of investors’ choices between impact and traditional funds remain underexplored. The study builds on insights about altruism, risk preferences, and nudging (e.g., visual cues and information framing) and references industry narratives about generational attitudes toward impact investing, as well as evidence on gender differences in altruism and investing behavior.

Design: An online, individual experiment implemented via oTree. Participants completed demographics, a prior-knowledge check, an effort task, and then made choices across eight investment scenarios. Monetary incentives were real: approximately 10% of participants were randomly selected for payout based on one randomly chosen decision. Sample: N=602 (541 non-experts; 61 experts). Gender: 367 female (341 non-experts, 26 experts), 235 male (200 non-experts, 35 experts). Participants spanned multiple countries, with many non-experts residing in Spain. Non-experts were recruited via non-probability sampling; experts via professional networks and referrals. Prior knowledge manipulation: Participants stated whether they had prior knowledge of impact investing. Those without prior knowledge had a 50% chance to view a 2-minute tutorial video (MBA, 2017) before proceeding. Effort task and endowment: Subjects counted the number of ‘1’s in a binary string; performance determined their investment budget. Correct solution rates: 90.3% experts, 94.4% non-experts. Decision scenarios (binary TIF vs. IIF choices):

• Q1: Baseline choice. TIF 5% vs. IIF 4% plus social impact (clean water in developing countries).
• Q2: Return trade-off. TIF 5% vs. IIF with 4%, 3%, 2%, or 1% return (with impact). Tests willingness to sacrifice financial return.
• Q3–Q4: Impact description framing. As Q1–Q2 but with additional concrete information on realized social impact.
• Q5–Q6: Risk framing. As Q1–Q2 but with return risk: TIF 90% chance of return vs. IIF 80% chance (no social impact for TIF; impact for IIF).
• Q7: Fiscal incentive. TIF 5% vs. IIF 4% plus 20% tax deduction on the invested amount.
• Q8: Visual aid. As impact description with added image and geographic context (Papua New Guinea water access) to evoke empathy without victimization. Analysis: Logistic regressions assessed the effects of framings and demographics on choosing IIF (vs. TIF). Models included scenario framings (impact description, risk, tax deduction, visual aid), multiple return options, and the return difference (Delta) between TIF and IIF. Demographic covariates: gender, expert status, age, education levels, prior knowledge (self-report), video displayed indicator, and Delta (when applicable).

• Baseline preference: In Q1, 85.2% chose IIF over TIF.
• Framing effects (Table 5):
  • Impact description increased IIF choice (p<0.05).
  • Visual aid (image plus details) further increased IIF choice (p<0.001).
  • Risk framing (IIF 80% vs. TIF 90% return chance) decreased IIF choice (p<0.001).
  • Tax deduction (20%) showed no significant overall effect.
• Return trade-off (Table 6): Across framings (no extra info; impact info; risk), larger negative Delta (IIF return lower than TIF by more) reduced IIF choice (all p<0.001).
• Demographics (Table 7):
  • Gender: Women were more likely than men to choose IIF in most scenarios, including under risk and with visual aid (p-values ranging from <0.01 to <0.001). No significant gender difference for tax deduction framing.
  • Expertise: Experts were more likely to choose IIF than non-experts in baseline and impact-information scenarios (p<0.001). Under risk, expert effect was not significant; with tax deduction, experts showed a positive response (p<0.05).
  • Age: Greater age associated with higher IIF choice in baseline (p<0.001) and under risk (p<0.001). Age effect not significant when additional impact information or visual aid was provided; under tax deduction, older subjects were more likely to choose IIF (p<0.01).
  • Education: No consistent significant effects; generally not a substantial predictor (one negative effect under risk for higher education: p<0.05).
  • Prior knowledge (self-reported): Not a significant predictor.
  • Video displayed: Viewing the tutorial video increased IIF choice in baseline (p<0.001), under risk (p<0.01), and especially with tax deduction (p<0.001); no effect with visual aid. Overall, information about social impact—especially accompanied by images—promotes IIF choices; perceived investment risk and larger return gaps discourage them. Women, experts, and older individuals show stronger inclinations toward impact options.

The higher propensity of experts to choose IIF likely reflects self-selection into social finance motivated by ethical principles. Contrary to popular narratives about younger generations driving impact investing, older participants exhibited greater willingness to invest for impact, potentially reflecting legacy motives or higher confidence/experience. Women’s greater inclination toward IIF aligns with literature on altruism and industry observations of values-aligned investing among women; notably, women’s preference persisted even under higher risk. Information interventions matter: providing concrete impact descriptions and especially visual aids increased IIF choices, and a brief educational video nudged participants toward impact investments. In contrast, tax incentives did not generally shift behavior—except among experts and older participants—suggesting limited efficacy of such policies for broad audiences. The results underscore the importance of awareness, communication strategies, and presentation framing in promoting impact investing and suggest that reliance on fiscal incentives alone may be insufficient. Persistent skepticism and cultural divides within impact investing ecosystems were also observed, highlighting the need for better cross-sector understanding and education.

This experiment with 602 participants identifies who is more likely to invest for impact and which levers influence such choices. Participants generally favored IIF, especially women, experts, and older individuals. Information about realized impact and visual aids significantly increased IIF selection, while higher perceived risk and larger return differentials reduced it. Tax incentives did not generally change behavior, though experts and older individuals were more responsive. Prior self-reported knowledge did not matter, but a short educational video nudged participants toward IIF. Main contributions include behavioral insights for designing communication, education, and policy strategies to mainstream impact investing. Future research should: (i) use broader and more diverse samples, particularly more experts; (ii) evaluate the effectiveness and design of fiscal incentives across countries; (iii) test public disclosure and reputational effects on choices; and (iv) integrate impact investing education into finance curricula and professional training.

The primary limitation is the small expert sample, reflecting recruitment challenges and time constraints. While non-experts were numerous, the non-probability sampling and sample composition may limit generalizability, and expert data are not publicly available to preserve anonymity.