Similar but yet different: individual cognitive traits and family contingencies as antecedents of intrapreneurship and self-employment

The study investigates how individuals choose among intrapreneurship, self-employment, and non-entrepreneurship. While many employees with entrepreneurial talent remain within organizations engaging in intrapreneurship, the determinants of choosing intrapreneurship versus self-employment remain unclear. Grounded in entrepreneurial action theory and conservation of resources (COR) theory, and incorporating a family embeddedness perspective, the paper asks: Do the same socio-cognitive traits (entrepreneurial alertness and entrepreneurial self-efficacy) similarly shape intrapreneurship and self-employment? How does their interaction influence choices? Which family contingencies (work–family conflict and household income) moderate these relationships? The study proposes that EA and ESE each promote both intrapreneurship and self-employment but more strongly the latter; that their interaction reduces intrapreneurship but increases self-employment; and that work–family conflict and household income differentially moderate these effects.

The paper reviews entrepreneurship and intrapreneurship literatures, distinguishing intrapreneurship (opportunity development within organizations) from self-employment and hybrid entrepreneurship. It critiques early assumptions that intrapreneurs and entrepreneurs share identical drivers, highlighting employee embeddedness and organizational constraints. Socio-cognitive traits from social cognitive theory—entrepreneurial alertness (EA) and entrepreneurial self-efficacy (ESE)—are linked to opportunity recognition and action. Entrepreneurial action theory differentiates third-person (opportunity exists) and first-person (opportunity for me) opportunities, suggesting EA relates to attention/identification while ESE relates to evaluation/action. COR theory frames entrepreneurial choice as resource investment versus loss avoidance. Family embeddedness research indicates families provide resources and constraints; however, prior work largely ignores family effects on intrapreneurship. The authors develop hypotheses: H1 and H2 posit positive effects of EA and ESE on both intrapreneurship and self-employment, stronger for self-employment; H3 posits a negative EA×ESE effect on intrapreneurship and positive on self-employment; H4 posits work–family conflict weakens these effects (less intrapreneurship negativity, less self-employment positivity); H5 posits household income strengthens these interaction effects (more negative for intrapreneurship, more positive for self-employment).

Two studies were conducted.

• Study 1: Secondary data from the Global Entrepreneurship Monitor (GEM) special surveys 2011–2019, merged with World Bank GDP per capita and ILO unemployment rates. After cleaning (ages 18–64, complete data, truncating top 0.1% of household size), N=704,515 individuals across 97 economies. Measures: Entrepreneurial choice (EC) coded as 0=non-entrepreneurship, 1=intrapreneurship (trying to start a new venture for employer), 2=self-employment (trying to start a business for oneself); EA (binary: perceives good business opportunities); ESE (binary: has knowledge/skills/experience to start a business); household size as proxy for work–family conflict; household income (tertiles: 1–3). Controls: gender, age and age^2, education, GDP per capita (log, lagged), unemployment rate (lagged), year dummies. Empirical strategy: Multinomial logistic regressions with non-entrepreneurship as baseline (and alternatively intrapreneurship); interaction terms EA×ESE; moderation models including Size and Income with two-way and three-way interactions; robust SEs clustered by economy; model fit via log pseudo-likelihood and Pseudo-R².
• Study 2: Extended replication survey in China to address Study 1 limitations. Data collected via community committees in 17 subdistricts across 5 cities; both paper and electronic formats; back-translation of scales. Final N=363 valid working-age respondents (45.7% female; mean age=39; 68.6% high school+; 84.6% employed; 5.2% with entrepreneurial experience). Measures: Intrapreneurship behavior (IB) via Gawke et al. (2019) venture behavior scale (7 items, 7-point); Self-employment (SE) behaviors via a 10-item checklist (yes/no); EA via 7-item short Tang et al. (2012) scale (7-point); ESE via Schjoedt & Craig (2017) 3-item scale (7-point); Work-to-family conflict (WFC) and Family-to-work conflict (FWC) via Netemeyer et al. (1996) scales (7-point); Household income in 11 brackets (¥30,000 increments up to >¥300,000). Controls: gender, age and age^2, education, entrepreneurial experience. CFA confirmed 5-factor structure (CFI=0.985, RMSEA=0.024, SRMR=0.046). Harman single-factor=21.8% variance. Empirical strategy: Two linear regression models (DV=IB and DV=SE), with centered interactions EA×ESE and three-way interactions with WFC, FWC, and Income; multicollinearity checks (mean VIF 1.116–1.246); model fit via adjusted R².

Study 1 (GEM, N=704,515):

• EA positively predicted intrapreneurship (RR=1.449, 95% CI [1.323, 1.586]) and self-employment (RR=2.475, 95% CI [2.231, 2.745]); ESE positively predicted intrapreneurship (RR=2.126, 95% CI [1.966, 2.299]) and self-employment (RR=5.202, 95% CI [4.712, 5.744]).
• Relative to intrapreneurship, EA (RR=1.708, 95% CI [1.590, 1.834]) and ESE (RR=2.447, 95% CI [2.248, 2.664]) increased the likelihood of self-employment (supporting stronger effects on self-employment: H1c, H2c).
• Interaction EA×ESE negatively associated with intrapreneurship (RR=0.777, 95% CI [0.703, 0.859]; H3a supported). For self-employment, EA×ESE coefficient was positive but not significant in logit (RR=1.048, p=0.426); marginal effects plots indicated a positive interaction (H3b supported).
• Moderation by household size (proxy WFC): Three-way EA×ESE×Size not significant for intrapreneurship (RR=1.001, p=0.947; H4a not supported). For self-employment, larger household size weakened the positive EA×ESE effect (RR=0.971, 95% CI [0.945, 0.999]; H4b supported).
• Moderation by household income: Higher income strengthened the negative interaction effect on intrapreneurship (EA×ESE×Income RR=0.805, 95% CI [0.747, 0.869]; H5a) and strengthened the positive interaction effect on self-employment (RR=1.145, 95% CI [1.055, 1.243]; H5b).

Study 2 (China survey, N=363):

• Intrapreneurship behavior: EA (β=0.090, 95% CI [0.019, 0.162]) and ESE (β=0.084, 95% CI [0.005, 0.162]) positively related (H1a, H2a). EA×ESE negative (β=-0.082, 95% CI [-0.137, -0.027]; H3a). Moderation: EA×ESE×FWC positive (p=0.049), indicating at high FWC, those high in EA and ESE show more intrapreneurship; EA×ESE×Income negative (β=-0.027, 95% CI [-0.043, -0.011]; H5a). EA×ESE×WFC not significant.
• Self-employment behavior: EA (β=0.175, 95% CI [0.047, 0.304]) and ESE (β=0.285, 95% CI [0.145, 0.425]) positively related (H1b, H2b). EA×ESE positive (β=0.102, 95% CI [0.003, 0.201]; H3b). Moderation: EA×ESE×WFC marginally positive (90% CI [0.003, 0.125]), suggesting high WFC can push high EA/ESE individuals toward self-employment; EA×ESE×FWC negative (β=-0.082, 95% CI [-0.162, -0.002]), and EA×ESE×Income marginally positive (p=0.025, 90% CI [0.001, 0.050]; H5b trend).

Overall: EA and ESE each increase both intrapreneurship and self-employment, more strongly the latter. Their interaction reduces intrapreneurship but increases self-employment. Family context matters: higher FWC steers high EA/ESE individuals toward intrapreneurship (and away from self-employment), while higher WFC and higher household income steer them toward self-employment.

Findings clarify that entrepreneurial choice is governed by both opportunity cognition and resource considerations. EA aids opportunity recognition; ESE facilitates action, and together they shift individuals toward self-employment when both are high—consistent with entrepreneurial action theory’s transition from third-person to first-person opportunities. Conversely, intrapreneurship becomes attractive when only one of the two traits is high or when family constraints (especially FWC) increase the perceived costs of independent venturing, aligning with COR theory’s loss-avoidance emphasis. The results show that family embeddedness nuances these processes: high FWC provides incentives to leverage organizational resources via intrapreneurship, while high WFC and greater household income increase tolerance for risk and the pursuit of autonomy through self-employment. This integrated view advances understanding of when similar socio-cognitive traits yield different entrepreneurial paths and underscores the importance of directionally distinct work–family conflicts in shaping entrepreneurial action.

The paper integrates intrapreneurship and self-employment into a unified entrepreneurial choice framework. It demonstrates that EA and ESE are common antecedents for both forms but interact to favor self-employment and disfavor intrapreneurship. Family contingencies play key moderating roles: FWC encourages intrapreneurship among high EA/ESE individuals, whereas WFC and household income push toward self-employment. The work advances entrepreneurial action theory by incorporating organizational contexts (intrapreneurship) as a mechanism connecting third- and first-person opportunities and extends COR and family embeddedness perspectives to entrepreneurial choice. Future research should employ richer, multi-item measures for EA/ESE, longitudinal or experimental designs to address causality and common method bias, disentangle WFC vs. FWC dynamics over time, consider additional socio-cognitive traits (e.g., risk preference, fear of failure), and analyze entrepreneurial choice in frameworks incorporating hybrid entrepreneurship.

Key limitations include: (1) Study 1 relied on single-item, binary measures of EA and ESE from GEM, threatening construct validity and precluding nuanced assessment; (2) Work–family conflict was proxied by household size in Study 1, conflating family demands with household human/social capital and not distinguishing WFC from FWC; (3) Predominantly self-reported, cross-sectional data raise concerns about common method bias and reverse causality; (4) Categorical outcomes complicate interaction inference in multinomial logit (addressed via plots but still a constraint); (5) GEM data lag (2011–2019) may affect timeliness; (6) Generalizability of Study 2 is limited to Chinese sub-samples; (7) Not all hypothesized moderations (e.g., WFC effects on intrapreneurship) were consistently supported across studies.