When Environmental Sustainability Meets Equity Crowdfunding: The Role of Green Ventures in Job Creation

Resource-Based View and Natural Resource-Based View

RBV (Barney, 1986, 1991; Lippman & Rumelt, 1982; Penrose, 1959; Wernerfelt, 1984) of the firm explains how managers use strategic resources and develop capabilities (Peteraf, 1993; Teece et al., 1997) to attain and sustain competitive advantage. Traditionally RBV emphasizes that strategic resources generate competitive advantage when they are valuable, rare, difficult to imitate, and effectively organized (Barney, 1991; Helfat et al., 2023; Nason & Wiklund, 2018). A capability, in contrast, is something a firm can perform that stems from resources and routines (Peteraf, 1993; Teece et al., 1997). Both resources and capabilities are continuously developed and improved within the firm over time (Helfat & Peteraf, 2003).

A complementary layer to RBV, NRBV accounts for constraints imposed by firms’ biophysical environment (Hart, 1995; Hart & Dowell, 2011). Rather than highlighting the importance of resources, the NRBV focuses on the development of three capabilities that can help firms gain competitive advantage: (a) Pollution prevention; (b) Product stewardship; (c) Sustainable development. In more recent developments of the NRBV, Hart and Dowell (2011) extended the set of capabilities to further include clean technology and base of the pyramid strategies. Menguc and Ozane (2005) argued that environmental orientation by itself can be viewed as a capability rooted in commitment to the natural environment, corporate social responsibility and entrepreneurship.

Traditionally RBV and NRBV are applied to large established companies with already developed capabilities and resources in place, or able to buy them in the factor markets. This makes RBV about the choice of how to optimize the use of resources and capabilities to grow the firm, which sets clear boundary conditions for which RBV and NRBV are directly applicable. The characteristics of entrepreneurial ventures call for a more nuanced application of RBV and NRBV (Alvarez & Busenitz, 2001; Shane & Venkataraman, 2000; Zahra, 2021). As young ventures suffer from the liability of newness (Stinchcombe, 1965) and/or smallness (Brüderl & Schussler, 1990), they are both time- and size-constrained in the development and acquisition of strategic resources (Katila & Shane, 2005) and capabilities (Arikan & McGahan, 2010). Therefore, entrepreneurial ventures require different resources and capabilities, or use them differently compared to larger, more established firms (Unger et al., 2011; Wiklund et al., 2009). Young ventures also face other obstacles apart from the lack of strategic resources and developed capabilities due to newness and smallness. In particular, information asymmetries between entrepreneurs and external resource providers further constrain resource acquisition (Brinckmann & Hoegl, 2011; Janney & Dess, 2006; Stuart et al., 1999).

In this study, we extend RBV and NRBV to an entrepreneurial setting where the lack of conventional strategic resources, developed capabilities, and information asymmetries make resource acquisition and capability development challenging. We conceptualize environmental orientation as an early-stage organizational capability that offers two complementary but theoretically distinctive mechanisms explaining resource acquisition in young ventures. First, ventures’ environmental orientation may increase the demand for human capital resources. The set of environmental capabilities discussed in Hart (1995), and Hart and Dowell (2011) are inherently knowledge intensive and require substantial investment in human capital to develop. The greater need for financial capital and human capital should lead to greater demand for both capital and labor in EOVs relative to NEOVs.

Second, from a supply-side perspective, the observable expression of environmental orientation through campaign communications, organizational identity, and hiring-related communication can convey credible signals to investors and prospective employees (Kleinert & Vismara, 2025; Mansouri & Momtaz, 2022; Vismara, 2019). Most studies integrating signaling theory into RBV try to explain why some firms may have greater resource access. These studies traditionally focus on reputation, an intangible strategic resource, that can provide competitive advantage through signaling in capital and factor markets (Boyd et al., 2010; Rindova et al., 2005). However, Boyd et al. (2010) highlighted evident boundary conditions. Young ventures lack the history and track record to accumulate reputational capital. Instead, distinct from viewing environmental orientation as a resource, we focus on environmental orientation as an early-stage organizational capability. This capability reflects the venture’s ability to recognize, prioritize, and organize around environmental opportunities and constraints. We argue that ventures’ environmental orientation is an important part of defining the organization’s identity, which could signal legitimacy to external parties such as prospective investors and employees (Lounsbury & Glynn, 2001). However, the signaling value of capabilities to external parties is dependent on whether the venture is organized enough to display it through communications, campaign disclosures, and hiring practices (Skaggs & Snow, 2004).

Role of Environmental Orientation in Financial Resource Acquisition under Asymmetric Information

In traditional RBV formulations financial resources do not qualify as a strategic resource as they fail to meet standard criteria (Nason & Wiklund, 2018; Zahra, 2021). Financial capital is instead viewed as a versatile resource with the advantage that it can be converted into other types of resources in the factor markets (Wiklund & Shepherd, 2005). As young ventures often fail to internally generate positive cash flows while simultaneously facing numerous profitable investment opportunities, access to financial resources is crucial for survival and growth (Coad et al., 2013; Cooper et al., 1994; Cosh et al., 2009; Holtz-Eakin, et al., 1994; Linder et al., 2020). At the same time, young ventures are disadvantaged in raising financial resources from conventional financing sources as they face liabilities of newness and smallness in addition to being opaque (Battisto et al., 2019; Beck et al., 2005; Carpenter & Petersen, 2002).

Given these constraints, opaque young ventures increasingly turn to ECF to overcome hurdles in financial resource acquisition (Cumming et al., 2021; Lazos, 2025; Walthoff-Borm et al., 2018a). ECF offers access to a wider investor base compared to conventional financing such as banks, venture capitalists or business angels, with investors often being less experienced than professional ones (Nitani et al., 2019; Schwienbacher, 2019; Shafi, 2021; Vismara, 2018, 2019). Thus, their investment and screening criteria likely differ from those of traditional investors (Hartzmark & Sussman, 2019). As a result, they may have different investment processes and objectives allowing for investments into ventures that would not appeal to professional investors. By contrast ECF may increase information asymmetry between ventures and investors. ECF proposals are coupled with lower levels of disclosure in the prospectus relative to traditional funding sources (Johan & Zhang, 2020) and thus a greater degree of information asymmetry between investors and ventures (Ahlers et al., 2015). Instead of traditional investor documentation, ventures seeking crowdfunding aim to use more soft information in investor pitches than traditional firms do through text descriptions, videos, pictures, and testimonials from others in the crowd (Johan & Zhang, 2022). Therefore, ECF investors make decisions based on other variables than pure financial ones (Shafi, 2021) such as emotions (Sitruk et al., 2025), and social values (Lehner, 2013). In opaque ECF environments, signaling therefore becomes an important factor in financial resource acquisition to separate the offering from others (Ahlers et al., 2015; Courtney et al., 2017; Mansouri & Momtaz, 2022; Scheaf et al., 2018; Spence, 1973).

To capture the attention of the ECF investors, ventures can convey their environmental orientation through their investor communications (Mansouri & Momtaz, 2022). Environmental orientation may function as a strategic capability displayed through campaign communications and can mitigate some information asymmetries by conveying the venture’s value for pro-environmental investors. The empirical evidence supports the claim as environmental orientation increases investors’ willingness to pay (Barber et al., 2021; Heeb et al., 2023; Hornuf et al., 2022; Mansouri & Momtaz, 2022). This indicates that environmentally oriented investors evaluating two identical ventures (with similar expected cash flows and risk) only differing in their environmental orientation would attach a greater value to the EOVs. Also from a venture perspective, consistent with NRBV, greater demand for human capital investments needed to further develop environmental capabilities may also in turn increase the demand for capital. Accordingly, our first hypothesis posits that environmental orientation increases the funding under the RBV’s and NRBV’s entrepreneurial extensions.

Role of Environmental Orientation in Financial Resource Allocation

Securing financial resources in the campaign stage is only the first hurdle in an entrepreneurial venture’s pursuit of survival and growth. Once the financial resources are acquired, the entrepreneurial team must decide in the post-campaign stage on the allocation and management of the resources across areas such as human capital, physical capital, or intangibles such as R&D and marketing. RBV and NRBV have typically emphasized the strategic value of resources and capabilities but paid less attention to how entrepreneurs’ orientations shape resource allocation (Deng et al., 2020; Zahra, 2021). In this study, we theorize that ventures’ environmental orientations impact financial resource allocation decisions through increased demand for human capital and reduced frictions in labor supply.

We argue that environmental orientation affects employee growth through two complementary reinforcing channels. First, on the demand-side NRBV highlights the greater need for human capital investments to develop capabilities (Hart, 1995; Hart & Dowell, 2011). If EOVs require more or higher-skilled labor, this should be reflected in their resource allocation decisions. In evidence for discrepancies between EOVs and NEOVs, Consoli et al. (2016) put forward that environmentally oriented firms allocate a greater proportion of human capital into their production process. However, we also know that industries have heterogenous production functions, where labor and capital inputs are different, irrespective of the firm being an EOV or NEOV. To better identify the demand for human capital relative to physical capital, we need to compare EOVs and NEOVs within the same industry. If EOVs hire more after the inclusion of industry and physical capital controls, it would point toward either a greater demand or supply of labor for EOVs relative to NEOVs.

Second, on the supply-side, environmental orientation may enlarge the applicant pool and improve access to talent. Labor market frictions make talent acquisition a key constraint for new ventures (Rynes & Barber, 1990). Conveying environmental orientation in communications can mitigate this constraint by enlarging the pool of applicants. Because individuals with pro-environmental preferences are more inclined to match with employers holding similar values (Greening & Turban, 1997, 2000). Thus, environmental orientation may translate into greater employee growth through increased labor supply.

Another important aspect is the amount raised in the ECF campaign, as we would expect firms that raise a greater amount in the campaign to hire more irrespective of environmental orientation. Conditional on EOVs obtaining a greater amount of funding, it is expected that EOVs will create more jobs. Hence, to better understand the allocation decision, we need to account for the financial resources raised in the campaign. By controlling for the amount raised in the campaign and proxies for the production function, we can better identify the impact of environmental orientation on employment, that is, to see if the demand and supply effects from the ventures’ environmental orientation predict a greater allocation to human capital.

The next hypothesis evaluates whether human capital allocations stem from demand-side considerations. From a demand-side perspective, consistent with NRBV, the allocation should be dependent on the greater need for human capital. Entrepreneurs disclose their environmental orientation and their prospective use of funds already in the campaign stage prior to knowing the success of the offering. As intent is formulated prior to knowing the campaign outcome, it is potentially less subject to endogeneity and distinct from realized job creation that is more likely to be endogenous to funding success. To evaluate whether hiring by EOVs is demand driven, we study whether EOVs are more likely to show an intent to allocate resources toward human capital in the prospectuses.

Next, we move from the intent in the campaign stage to the post-campaign stage hiring outcomes to verify whether environmental orientation can convey signals to attract talent in TLMs. In a world absent of labor market frictions, venture location and industry should not matter for employee growth. However, Manning and Petrongolo (2017) reported that labor markets are local, with large regional differences in labor supply coupled with an imperfect cross-region labor mobility. Furthermore, rapid technological shifts have created a mismatch between labor demand and supply at the industry level (Thurik, 2003). This creates heterogeneous supply and competition for human capital across regions and industries. In tighter labor markets, ECF-financed firms may face competition for talent from large incumbent firms and venture capital backed companies which can make hiring difficult. However, even if the competition for labor is tough in TLMs, the environmental signal may help to increase the supply of talent. Hence, we argue that EOVs find it easier to attract talent even in tight regional and sectoral labor markets characterized by low unemployment and high competition for human capital.

Figure 1 summarizes the theoretical framework. We conceptualize environmental orientation as an early-stage organizational capability that affects campaign-stage fundraising and post-campaign employee growth. First, environmental orientation may increase the amount of capital raised in the campaign (H1). Second, conditional on funding EOVs may exhibit higher employee growth (H2). We examine two complementary mechanisms for this relationship: a demand-side mechanism based on the need to develop environmental capabilities (H3a), and a supply-side mechanism based on the signaling value of environmental orientation in labor markets (H3b).

OPEN IN VIEWER

Figure 1.

Theoretical link between the hypotheses.

Descriptive Statistics

Figure 2 shows employee growth over time relative to the ECF campaign year t. We have indexed the graph at 100 at t−1 to facilitate comparison between EOVs and NEOVs. We can observe that 2 years after the campaign EOVs have a net growth in headcount of roughly 5× and 3× from t−1 by using the Beauhurst and text-based classifications. NEOVs observe a more modest growth for both measures (2× and 2.25×). Worth noting is that the employee growth between EOVs and NEOVs is similar between t−1 and t, especially by using the text-based measure. We observe similar differences using our main employee growth measures from t to t+1.

OPEN IN VIEWER

Figure 2.

Employee growth over time.

The descriptive statistics in Panel A of Table 1 show the mean, standard deviation, minimum, and maximum of the variables used in the study. Our dataset covers 659 successful ECF campaigns. The average employee growth conditional on obtaining financing is 40.56% or 2.26 employees. We are able to identify 47 sustainable ventures or 7.1% of the sample using the Beauhurst classification and 87% or 13.2% using our text-based classification. This is slightly lower than the numbers found in Vismara (2019) and Vismara and Wirtz (2025) that report an environmental sustainability fraction of around 20%. The higher numbers are potentially due to their different classification methods; we discuss the implications in the limitations chapter. On average, our sample firms raise £740,000 in the campaign. This makes our campaign amounts larger than those in comparable studies on US data where Cumming et al. (2025) found an average campaign size of $399,000. In UK samples, Vismara (2019) reported an average amount of £251,030, and Vismara and Wirtz (2025) reported £305,000. Those studies deploy data from unsuccessful offerings as well, that may yield lower amount values. The average number of employees prior to the successful ECF campaign amounts to 8.94. The sample ventures have on average 1.74 times total assets in liabilities, this is due to the large fraction of firms with negative equity. The firms hold on average 28.80% of their assets in cash, and 21.77% in fixed assets.

OPEN IN VIEWER

Table 1.

Summary Statistics.

Variables	N	Mean	SD	Min		Max
Panel A
Employee growth (in %)	547	40.557	83.645	−100.000		400.000
ln(Amount)	659	12.957	0.971	10.940		15.617
Intent to recruit	481	0.366	0.482	0.000		1.000
EOV	659	0.071	0.258	0.000		1.000
EOV2	659	0.132	0.339	0.000		1.000
Amount (in m)	659	0.744	1.191	0.026		15.000
Team size	659	2.753	1.481	1.000		7.000
Firm age	659	7.042	0.755	4.990		8.703
Diversification	659	1.310	0.734	1.000		4.000
Firm size	659	8.994	19.864	0.000		141.000
Leverage	437	1.737	5.874	0.000		49.012
Cash	437	0.288	0.342	0.000		1.000
Tangibility	437	0.218	0.294	0.000		0.998
Tight labor market	603	0.000	1.213	−5.073		3.491
Competing offerings (per year)	659	501.206	55.628	429.000		569.000
Successful offers (per quarter)	659	118.023	20.930	88.000		149.000
	EOV	NEOV	Difference	EOV2	NEOV2	Difference
Panel B
Employee growth (in %)	85.735	36.896	48.839***	63.053	36.759	26.294**
ln(Amount)	13.145	12.943	0.201	13.145	12.929	0.216**
Intent to recruit	0.447	0.359	0.088	0.369	0.365	0.004
Team size	2.702	2.757	−0.054	2.724	2.757	−0.033
Firm age	7.033	7.042	−0.009	7.038	7.042	−0.004
Diversification	1.426	1.301	0.125	1.437	1.290	0.147
Firm size	5.617	9.253	−3.636	5.908	9.463	−3.555***
Leverage	0.974	1.803	−0.829	1.725	1.739	−0.014
Cash	0.250	0.292	−0.042	0.274	0.291	−0.016
Tangibility	0.203	0.219	−0.016	0.157	0.228	−0.071**
Tight labor market	0.131	−0.010	0.140	−0.330	0.052	−0.382***
Competing offerings (per year)	502.234	501.127	1.107	518.713	498.544	20.169***
Successful offers (per quarter)	113.617	118.361	−4.744	117.207	118.147	−0.940
Observations	47	612		87	572

Note. This table reports summary statistics of variables employed in this study. Panel A reports the mean, standard deviation, minimum, and maximum values of the variables. Panel B includes means and mean difference tests between EOVs and NEOVs. We use the Beauhurst definition to create EOV, while we use a list of environmental sustainability related keywords to create EOV2. The sample period spans from January 2019 to December 2022 using a sample of successful Crowdcube and Seedrs campaigns. All variables are defined in Appendix A. EOV = environmentally oriented venture; NEOV = non-environmentally oriented venture; SD = standard deviation.

*

p ≤ .1. **p < .05. ***p < .01.

In Panel B of Table 1, we compare the means of EOVs to those of NEOVs. Our univariate findings on ln(Amount) partly lend support to Hypothesis 1. Based on our second classification (EOV2), EOVs raise significantly more capital relative to NEOVs (p < .05). However, this is not surprising as prior research reports similar evidence in other samples (see, Mansouri & Momtaz, 2022). New to the literature, we test Hypothesis 2, analyzing whether EOVs create more jobs. In support of Hypothesis 2, we report that the average employee growth among EOVs is 85.74% and NEOVs is 36.90% (p < .01); we report similar findings using the text-based classification. Even though EOVs are smaller relative to NEOVs at the capital raising stage, they manage to create more jobs in absolute terms (3.9 vs. 2.1). The EOVs and NEOVs are similar in terms of diversification, team size, firm age, liabilities, and cash holdings, but NEOVs have a greater proportion of fixed assets based on the second classification. Supplemental Table IA5 further shows pairwise correlations between variables used in the regressions. Interestingly, we show a significant pairwise positive correlation between the EOV measures and employee growth (p < .05).

EOVs and Capital Raising Outcomes

Our first regression aims to verify Hypothesis 1, that EOVs are able to raise more funding by displaying their environmental orientation to prospective investors. To test Hypothesis 1, we estimate the following regression:

ln ( Amount ) i = α + β 1 × * EO V i + X i δ ′ + FE

(1)

If EOVs raise more capital conditional on firm characteristics, we expect β 1 to be positive and statistically different from zero. X i is a matrix of control variables including Diversification, Team size, Firm age, Firm size, Leverage, Cash, and Tangibility. FE are fixed effects accounting for year, industry, region, and platform. We include year fixed effects as interest in both ECF and EOVs has increased over time, which warrants inclusion of year fixed effects. The year fixed effects also control for the COVID years. We further introduce industry fixed effects based on the UK Companies House classification from A to U to capture cross-industry differences in supply and demand for capital. We also include region ⁵ and platform fixed effects, as Cumming et al. (2025) reported important cross-platform heterogeneity in sustainability and survival, even though this is less of an issue on UK data compared to United States.

Column (1) of Table 2 shows estimates of model (1) including all covariates except financials. In support of Hypothesis 1, the coefficient estimate of 0.20 indicates that EOVs raise roughly 22.20% more capital in their campaigns relative to NEOVs (p = .02). ⁶ After including accounting variables, the number of observations drops from 656 to 433 in Column (2) due to data availability. However, the results remain similar to those in Column (1) (β = .25; p = .03). Next, we introduce our second environmental orientation measure (EOV2) in Columns (3) and (4), which classifies a greater number of firms as environmentally oriented. The coefficient estimates increase relative to those using the Beauhurst EOV measure to 0.25 (p = .01) and 0.30 (p = .01) in Columns (3) and (4), respectively. The economic magnitude remains large as EOVs raise 28.35% to 34.70% more capital in their ECF campaign. Hence, our results show consistently that EOVs raise more financial resources in their campaigns relative to NEOVs. We further estimate models using a different set of fixed effects in Supplemental Table IA7 to ensure that our findings are not driven by modeling choices. Among the control variables, Team size, Firm age, Firm size, and Cash are positively related to ln(Amount), while Leverage is negatively related.

OPEN IN VIEWER

Table 2.

Environmental Orientation and Amount Raised.

	(1)	(2)	(3)	(4)
Variables	ln(Amount)	ln(Amount)	ln(Amount)	ln(Amount)
EOV	0.2005**(2.701)	0.2453**(2.431)
EOV2			0.2496***(4.374)	0.2979***(3.271)
Team size	0.1653***(11.306)	0.1635***(6.448)	0.1646***(11.196)	0.1631***(6.237)
Firm age	0.2066***(5.594)	0.2011***(3.314)	0.2072***(5.891)	0.1998***(3.163)
Diversification	0.0430(1.561)	0.0012(0.034)	0.0396 (1.472)	−0.0032(−0.101)
Firm size	0.0162***(19.134)	0.0210***(4.370)	0.0162***(18.857)	0.0210***(4.379)
Leverage		−0.0166***(−4.249)		−0.0170***(−4.128)
Cash		0.4096***(5.166)		0.4083***(4.576)
Tangibility		−0.0259(−0.350)		−0.0062(−0.078)
Constant	10.8322***(45.089)	10.8164***(27.356)	10.8153***(47.505)	10.8049***(26.429)
Observations	656	433	656	433
R 2	0.394	0.386	0.399	0.393
Year FE	Y	Y	Y	Y
Industry FE	Y	Y	Y	Y
Platform FE	Y	Y	Y	Y
Region FE	Y	Y	Y	Y

Note. This table reports the amount raised in crowdfunding campaigns by EOVs relative to NEOVs. The dependent variable is the natural logarithm of total amount raised in equity crowdfunding [ln(Amount)]. We use the Beauhurst definition to create EOV, while we use a list of environmental sustainability related keywords to create EOV2. Heteroscedasticity robust t-stats clustered at the region level are reported in parentheses. All variables are defined in Appendix A. EOV = environmentally oriented venture; NEOV = non-environmentally oriented venture; FE = fixed effects.

*

p ≤ .1. **p < .05. ***p < .01.

Main Results: EOVs and Job Creation

We evaluate the main H2 by testing whether EOVs experience a greater post-campaign employee growth in excess of what is expected from the financial resources raised. To test H2, we estimate the following regression:

Employee growt h i = α + β 1 × * EO V i + β 2 × * ln ( Amount ) i + X i δ ′ + FE

(2)

If EOVs create more jobs conditional on funding and firm characteristics, we expect β 1 to be positive and statistically different from zero. A pure financial resource effect (relaxation of financial constraints) would instead render ( β 1 ) insignificant, leaving a statistically significant ( β 2 ) to fully explain the employee growth irrespective of environmental orientation. However, if EOV remains significant after the inclusion of ln(Amount), EOVs can create jobs beyond what is expected from the financial resource addition, which would lend support to the idea that EOVs find it easier to hire (supply-side) or have a greater labor demand (demand-side). By controlling for ln(Amount), we effectively compare the employee growth of EOVs and NEOVs that receive similar capital injections. In addition to ln(Amount), the tests include similar controls and fixed effects as in specification (Equation 1) above. We again include year fixed effects to control for macro-economic variations. This is important as our sample period spans over the COVID years which experienced large swings in labor supply and demand along with government policies. We include industry fixed effects to compare firms within the same industry, as sector-wide trends in employee growth and capital raising could potentially drive our results. Manning and Petrongolo (2017) showed that labor supply to a large extent is regional with non-significant cross-region mobility. To control for differences in labor supply, we include region fixed effects, allowing for a comparison between ventures that compete in similar local talent supply pools.

Column (1) of Table 3 shows that EOVs are linked to higher employee growth relative to NEOVs (β = 47.78; p = .04). The economic effect is large, EOVs create 47.78% more jobs the year after receiving ECF capital injections. Column (2) includes accounting controls making the sample size decrease from 545 to 357, due to data availability. The effect size increases following the inclusion of accounting variables in Column (2) (β = 61.95; p = .02); EOVs create 61.95% more jobs in the post-campaign stage. Next, we estimate similar models by including the text-based measure of EOV (EOV2). Also, EOV2 remains a significant predictor for post-campaign employee growth in Column (3) (β = 18.06; p = .08) and Column (4) (β = 28.58; p = .04). However, the economic magnitude reduces to 18.06% and 28.58% for Columns (3) and (4), respectively. Among the control variables, we report that Employee growth is negatively related to Firm age. Leverage and the level of cash holdings are positively related to Employee growth. ln(Amount) is statistically significant in three out of four specifications suggesting that both the financial resources obtained in the ECF campaign in addition to the EOVs’ supply and demand for labor are important determinants for employee growth. We further estimate the models using different types of fixed effects in Supplemental Table IA8 and find similar results. Hence, the evidence is consistent with Hypothesis 2: EOVs are linked to greater employee growth. By controlling for the amount raised, we show that EOVs hire in excess of what can be expected from the capital injection alone.

OPEN IN VIEWER

Table 3.

Employee Growth.

	(1)	(2)	(3)	(4)
Variables	Employee growth	Employee growth	Employee growth	Employee growth
EOV	47.7752**(2.297)	61.9484**(2.812)
EOV2			18.0552*(1.838)	28.5821**(2.187)
ln(Amount)	9.3424***(3.655)	7.2424(1.632)	9.4685***(3.893)	7.1517*(2.116)
Team size	−1.8061(−0.866)	2.6676(0.824)	−2.1465(−1.078)	2.0243(0.694)
Firm age	−21.4009***(−6.599)	−22.4836***(−3.357)	−21.5168***(−6.600)	−22.7630***(−3.476)
Diversification	5.1302(0.984)	3.8757(0.452)	5.2754(1.059)	3.8204(0.464)
Firm size	−0.1156(−0.817)	−0.1653(−0.398)	−0.1319(−0.856)	−0.1956(−0.446)
Leverage		1.7100***(8.838)		1.5583***(6.690)
Cash		35.1162***(3.006)		33.8522**(2.793)
Tangibility		12.2952(0.901)		11.4187(0.791)
Constant	66.1408(1.448)	73.9965(0.931)	67.2375(1.502)	80.6105(1.237)
Observations	545	357	545	357
R 2-squared	0.157	0.195	0.143	0.176
Year FE	Y	Y	Y	Y
Industry FE	Y	Y	Y	Y
Platform FE	Y	Y	Y	Y
Region FE	Y	Y	Y	Y

Note. This table reports the employee growth in EOVs relative to NEOVs after controlling for the amount raised in the campaign. The dependent variable is employee growth after the equity crowdfunding offering date calculated as the number of employees (t+1)/number of employees (t)−1, where t is the first report following the campaign end. We use the Beauhurst definition to create EOV, while we use a list of environmental sustainability related keywords to create EOV2. Heteroscedasticity robust t-stats clustered at the region level are reported in parentheses. All variables are defined in Appendix A. EOV = environmentally oriented venture; NEOV = non-environmentally oriented venture; FE = fixed effects.

*

p ≤ .1. **p < .05. ***p < .01.

In addition, we use two tests to account for potential endogeneity in our base case specifications: the 2SLS and the control function methods. ⁷ One potential source of endogeneity could stem from simultaneity between ln(Amount) and Employee growth. This could derive from firms with greater hiring capacity may also attract more funding, irrespective of environmental orientation. In the first stage, we estimate a similar model as (1) above on ln(Amount), with the inclusion of successful competing offers in the campaign. Our findings suggest that the number of successful offers is positively related to ln(Amount), with F-stats above 10 in all first-stage regressions (see Columns (1) and (4) Table 4). This is consistent with Cerpentier et al. (2022) who argued and showed that successful campaigns raise a greater amount of funding during hot issue markets, that is, during periods of elevated issuance activity. We further argue that it is not likely that the number of successful offers impacts employee growth in the second stage through other channels than the instrumented ln(Amount). Hence, the excluded exogenous first-stage variable likely satisfies both the inclusion and exclusion. For the 2SLS model, we obtain the fitted values from the first stage and include them in the second-stage estimations. Our findings in Columns (2) and (5) of Table 4 show that both EOV measures remain statistically significant predictors after accounting for potential endogeneity between ln(Amount) and Employee growth. This shows that the EOV effect on Employee growth also persists in the second stage after accounting for simultaneity between ln(Amount) and Employee growth.

OPEN IN VIEWER

Table 4.

Two-stage Least Squares and Control Function Estimations.

	(1)	(2)	(3)	(4)	(5)	(6)
	1st Stage	2nd Stage	2nd Stage	1st Stage	2nd Stage	2nd Stage
Variables	ln(Amount)	Employee growth	Employee growth	ln(Amount)	Employee growth	Employee growth
EOV	0.2771**(2.557)	77.6362***(3.833)	77.6362***(3.891)
EOV2				0.3466**(2.875)	46.3615**(2.247)	46.3615**(2.217)
Team size	0.1646***(4.863)	12.3397*(2.062)	12.3397*(2.019)	0.1634***(4.817)	10.7743*(1.997)	10.7743*(1.968)
Firm age	0.1592**(2.716)	−12.3936(−1.230)	−12.3936(−1.233)	0.1587**(2.570)	−13.5467(−1.449)	−13.5467(−1.459)
Diversification	0.0215(0.448)	5.2903(0.568)	5.2903(0.585)	0.0135(0.300)	4.7115(0.541)	4.7115(0.558)
Firm size	0.0212***(4.350)	1.0632(1.161)	1.0632(1.126)	0.0214***(4.324)	0.9318(1.087)	0.9318(1.056)
Leverage	−0.0309***(−8.073)	−0.1366(−0.116)	−0.1366(−0.114)	−0.0319***(−8.379)	−0.1817(−0.176)	−0.1817(−0.172)
Cash	0.4964***(5.052)	63.9507***(3.267)	63.9507***(3.207)	0.4989***(4.807)	60.2223***(3.008)	60.2223***(3.008)
Tangibility	−0.1118(−0.780)	5.9469(0.453)	5.9469(0.454)	−0.1005(−0.682)	6.2280(0.455)	6.2280(0.458)
Successful offers	0.0040***(3.189)			0.0043***(3.612)
Predicted amount		−51.7428(−1.486)			−46.5992(−1.619)
ln(Amount)			−51.7428(−1.476)			−46.5992(−1.613)
Control function residuals			59.5361(1.581)			54.3458*(1.820)
Constant	10.6101***(26.011)	723.1569*(1.939)	723.1569*(1.925)	10.5553***(25.849)	671.0055**(2.237)	671.0055**(2.222)
F-stats	10.17			13.05
Observations	357	357	357	357	357	357
R 2	0.416	0.193	0.197	0.426	0.175	0.178
Year FE	Y	Y	Y	Y	Y	Y
Industry FE	Y	Y	Y	Y	Y	Y
Platform FE	Y	Y	Y	Y	Y	Y
Region FE	Y	Y	Y	Y	Y	Y

Note. We use the number of successful offers to predict ln(Amount) in Column (1) and use the predicted amount and residuals in Columns (2) and (3), respectively to solve the simultaneity problem between ln(Amount) and employee growth. We repeat the same step with EOV2 in Columns (4) to (6). Heteroscedasticity robust t-stats clustered at the region level are reported in parentheses. All variables are defined in Appendix A. EOV = environmentally oriented venture; FE = fixed effects.

*

p ≤ .1. **p < .05. ***p < .01.

The control function method involves including the residuals from the first stage in the second-stage regressions; the residuals control for potential endogeneity stemming from simultaneity between ln(Amount) and Employee growth. The control function method yields identical estimates as the 2SLS with linear outcome variables, but with the advantage that it directly tests for potential endogeneity between ln(Amount) and Employee growth. Statistically significant first-stage residuals in the second-stage estimation indicate potential endogeneity problems. Our findings in Columns (3) and (6) of Table 4 suggest ln(Amount) and Employee growth are potentially endogenous in one of the specifications prior to the correction.

In auxiliary tests (Supplemental Table IA9), we further confirm that we observe excess hiring over different levels of ln(Amount) by entropy balancing EOVs and NEOVs based on the amount raised. By conducting this test, we ensure that the effect is consistent all over the capital raising distribution, that is, we observe excess hiring. Our results show that EOV remains a significant predictor of employee growth after balancing the sample. We further conduct sample splits in Columns (3) to (6) in Supplemental Table IA9 to show that the EOV effect on employee growth exists both for small and large capital injections. Our findings support H2.

Supply or Demand Considerations

The next set of tests aims to distinguish between supply- and demand-driven employee growth. From a demand-side perspective, human capital allocation is dependent on the greater structural demand of EOVs face as they develop environmental capabilities. To test for a demand-driven allocation (H3a), we make use of the fact that entrepreneurs disclose their environmental orientation and their prospective use of funds already in the campaign stage prior to knowing the success of the offering. As their intention to recruit is formulated already prior to knowing the campaign outcome, it is likely less subject to endogeneity and distinct from realized job creation that is more likely to be endogenous to funding success. ⁸ We estimate regression specification (Equation 3) below to test whether EOVs are more likely to show an intent to allocate resources toward human capital in the prospectus.

Intent to recrui t i = α + β 1 × * EOV + X i δ ′ + FE

(3)

If EOVs are more likely to show intent to recruit, we expect β 1 in the model above to be positive. Our findings in Columns (1) to (4) in Table 5 do not suggest that EOVs are more likely to disclose their intent to recruit relative to NEOVs (lowest p-value is 0.364). Hence, our estimations do not lend support to excess employee growth stemming from demand-side explanations.

OPEN IN VIEWER

Table 5.

Demand- and Supply-Side Considerations.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Variables	Intent to recruit	Intent to recruit	Intent to recruit	Intent to recruit	Employee growth	Employee growth	Employee growth	Employee growth
EOV	0.0615(0.627)	0.1039(0.938)			42.2170(1.664)	53.0951*(2.169)
EOV2			−0.0133(−0.179)	0.0402(0.491)			22.3044*(2.151)	39.7227*(2.231)
Tight labor market					−3.5529(−0.867)	8.6236(1.546)	−8.5061(−1.786)	1.9314(0.202)
EOV × Tight labor market					37.5203*(1.923)	36.8712*(1.936)
EOV2 × Tight labor market							16.3566(1.199)	26.7881(1.401)
Team size	−0.0168(−1.111)	−0.0260(−1.210)	−0.0171(−1.129)	−0.0256(−1.137)	−1.4602(−0.642)	3.8354(1.071)	−1.8260(−0.850)	3.7286(1.237)
Firm age	0.0182(0.610)	0.0389(1.325)	0.0188(0.651)	0.0366(1.276)	−19.4411***(−10.163)	−18.0048*(−1.933)	−20.2161***(−10.165)	−18.1798*(−2.002)
Diversification	0.0408(1.446)	0.0104(0.299)	0.0419(1.519)	0.0099(0.289)	5.4831(0.944)	4.5170(0.464)	5.3668(1.001)	3.5370(0.397)
Firm size	−0.0006(−1.216)	−0.0012(−1.075)	−0.0007(−1.292)	−0.0013(−1.158)	−0.1368(−0.769)	−0.3416(−0.521)	−0.1236(−0.632)	−0.3359(−0.484)
Leverage		0.0040***(4.594)		0.0039***(4.235)		1.6533***(8.061)		1.5952***(6.894)
Cash		0.1557(1.548)		0.1548(1.549)		37.4865**(2.673)		35.9643**(2.665)
Tangibility		−0.0584(−0.810)		−0.0564(−0.796)		17.3278(1.514)		13.6606(1.010)
ln(Amount)					9.9485**(2.753)	7.3657(1.118)	9.2458**(2.884)	6.5923(1.201)
Constant	0.2330(1.001)	0.1124(0.561)	0.2348(1.032)	0.1313(0.643)	42.3153(0.801)	37.8792(0.407)	58.6127(1.306)	51.1425(0.653)
Observations	479	311	479	311	495	319	495	319
R 2	0.2017	0.227	0.201	0.224	0.145	0.174	0.129	0.169
Year FE	Y	Y	Y	Y	Y	Y	Y	Y
Industry FE	Y	Y	Y	Y	Y	Y	Y	Y
Platform FE	Y	Y	Y	Y	Y	Y	Y	Y
Region FE	Y	Y	Y	Y	Y	Y	Y	Y

Note. This table examines the demand- and supply-side channels of EOVs affecting employee growth. Heteroscedasticity robust t-stats clustered at the region level are reported in parentheses. All variables are defined in Appendix A. EOV = environmentally oriented venture; FE = fixed effects.

*

p ≤ .1. **p < .05. ***p < .01.

Next, we test if environmental orientation can increase the supply pool of potential candidates and thus facilitate job creation. To test for a potential supply-driven explanation, we consider the tightness of the labor market. The greater competition for human capital in TLMs should put small ECF-funded firms at a disadvantage in hiring high-quality talent. However, even if the competition for labor is tough in TLMs, the environmental orientation signal may increase the supply of talent. Hence, we argue that EOVs find it easier to attract talent even in tight regional and sectoral labor markets characterized by low unemployment and high competition for human capital. We use our composite PCA-based index to capture labor market tightness, which includes regional unemployment, industry unemployment, and annual job ads scaled by employment at the industry- and city-level to capture labor market competition. We estimate regression specification (4) below to capture whether EOVs are more able to recruit in a TLM.

Employee growt h i = α + β 1 × * EOV + β 2 × * TLM + β 3 × * EOV × * TLM + X i δ ′ + FE

(4)

In our tests, we are mainly interested in the β 3 coefficient capturing the interaction effect between EOV and TLM on employee growth, where a positive coefficient would indicate that EOVs are more able to hire in TLMs. The interaction coefficient is positive in all specifications (Columns (5)–(8) in Table 5). However, β 3 only reaches conventional levels of statistical significance in the models using the Beauhurst classification of environmental orientation. Even though we do not find statistically significant results in all specifications, the sign of β 3 is positive in all models. We further run additional tests, splitting the TLM proxy at the median to avoid the impact of extreme observations. Our findings in Supplemental Table IA10 show a statistically significant relation between the interaction term and employee growth in three out of four specifications. We cannot draw too strong conclusions from our mechanism tests and should view the findings as suggestive rather than conclusive, as not all tests provide statistically significant interaction term coefficients, but our findings point toward a supply-side rather than a demand-side explanation.

Robustness Tests

We conduct a wide range of robustness tests to ensure the validity of our findings. We first aim to confirm that our findings are not driven by differences between EOVs and NEOVs. Our prior tests of the environmental orientation effect on capital raising and employee growth introduce two types of econometric problems. First, EOVs might be significantly different from NEOVs which impact the amount raised and the subsequent employee growth. To adjust for observable differences between EOVs and NEOVs, we employ the entropy balancing method (Hainmueller, 2012). Second, ventures receiving funding might be significantly different from ventures that do not. Our research question is based on how firms simultaneously can use environmental orientation to signal in capital and labor markets, and the net effect of EOV on employment after funding is received. To ensure that our results are not solely driven by funding success, we control for selection at the capital raising stage. To control for campaign outcome selection, we use a Heckman technique by modeling the probability of campaign success with the number of competing offerings along with several covariates as in Signori and Vismara (2018). ⁹ Third, we apply the Oster (2019) and Masten and Poirier (2026) coefficient stability tests to evaluate the severity of a potential omitted variable problem. Further robustness tests include alternative measures of employee growth, sample split on size, as well as tests on asset growth and an alternative EOV measure.

First, we consider that EOVs and NEOVs might be fundamentally different in terms of firm characteristics, by entropy balancing the sample based on (Diversification, Team size, Firm age, Firm size, Leverage, Cash, and Tangibility). ¹⁰ The estimations in Table 6 show that EOVs raise more capital and hire more personnel also after accounting for differences between EOVs and NEOVs using entropy balancing.

OPEN IN VIEWER

Table 6.

Robustness: Entropy Balanced Sample.

	(1)	(2)	(3)	(4)
Variables	ln(Amount)	Employee growth	ln(Amount)	Employee growth
EOV	0.2669***(3.292)	97.6547***(4.774)
EOV2			0.2742**(2.883)	28.2109*(2.107)
ln(Amount)		−27.1745**(−2.782)		−1.8356(−0.421)
Team size	0.1358***(8.820)	26.8136***(4.209)	0.1230***(3.559)	12.2252***(5.825)
Firm age	0.3875***(5.777)	11.6132(0.963)	0.1139(0.859)	−12.9207(−1.189)
Diversification	−0.0062(−0.094)	7.3483(0.802)	−0.0057(−0.112)	12.5362(1.017)
Firm size	0.0154(1.707)	−4.6358***(−5.191)	0.0343*(1.766)	−3.6660***(−4.090)
Leverage	−0.0691(−1.683)	5.4394(1.251)	−0.0129***(−3.231)	1.1077(0.993)
Cash	0.3974**(2.847)	78.0234**(2.922)	0.3945**(2.342)	30.3910*(2.102)
Tangibility	0.1905*(2.006)	15.0293(0.546)	0.2546(1.751)	10.0880(0.224)
Constant	9.6906***(22.936)	212.3135(1.647)	11.4279***(13.490)	112.3479(0.913)
Observations	433	357	433	357
R 2	0.534	0.498	0.389	0.345
Year FE	Y	Y	Y	Y
Industry FE	Y	Y	Y	Y
Platform FE	Y	Y	Y	Y
Region FE	Y	Y	Y	Y

Note. This table uses an entropy balanced sample to adjust for observable differences between EOVs and NEOVs. The mean, variance and skewness of firm characteristics before and after weighting are in Supplemental Table IA11. Heteroscedasticity robust t-stats clustered at the region level are reported in parentheses. All variables are defined in Appendix A. EOV = environmentally oriented venture; FE = fixed effects.

*

p ≤ .1. **p < .05. ***p < .01.

Next, we consider the role of selection in the campaign stage, where firms obtaining financing could be different from failed campaigns. We model the first stage as in Signori and Vismara (2018), by using the number of competing offerings (both successful and unsuccessful) at the time of the campaign to instrument for campaign success (Column (1) of Table 7) along with Team size, Firm age, and Diversification. Signori and Vismara (2018) and Coakley et al. (2025) found that competing offers negatively predict campaign success (whether the campaign is completed). Correia et al. (2025) did not find any impact on funding success (if firms reach their target amount) during the early days of ECF (time-period 2015–2018). Our sample starts in 2019 and continues during years of significant developments in the ECF market. Consistent with Signori and Vismara (2018), we find a negative and significant effect of competing bids on the success likelihood. We report similar outcomes as in prior estimations, EOV has a positive impact on ln(Amount) and Employee growth in Columns (2) to (5). In line with prior studies, we do not report a significant Inverse Mills ratio in any of the second-stage models. ¹¹ However, we acknowledge that this correction of selection bias in campaign outcomes does not account for potential selection effects due to rejection from other funding sources such as venture capital firms and business angels (see, Walthoff-Borm et al., 2018a). However, such a self-selection toward low-quality firms seeking crowdfunding should bias the effect of crowdfunding on employment downward rather than upward.

OPEN IN VIEWER

Table 7.

Accounting for Campaign-Stage Selection.

	(1)	(2)	(3)	(4)	(5)
Variables	Success	ln(Amount)	Employee growth	ln(Amount)	Employee growth
EOV		0.2453**(2.400)	61.9717**(2.869)
EOV2				0.3008***(3.332)	27.7639**(2.195)
ln(Amount)			7.1396(1.627)		7.1364*(2.122)
Team size	0.0582***(5.250)	0.1591(1.306)	14.0602(0.922)	0.1294(1.104)	10.3369(0.720)
Firm age	−0.1047***(−6.188)	0.2091(0.986)	−42.9913*(−1.764)	0.2606(1.334)	−37.7724(−1.669)
Diversification	0.2388***(23.751)	−0.0144(−0.040)	43.8948(0.695)	−0.1220(−0.362)	33.1040(0.570)
Firm size		0.0210***(4.329)	−0.1599(−0.376)	0.0210***(4.328)	−0.1940(−0.436)
Leverage		−0.0166***(−4.224)	1.6629***(7.563)	−0.0170***(−4.220)	1.5275***(5.965)
Cash		0.4099***(4.962)	34.0834***(2.995)	0.4108***(4.511)	33.0390**(2.790)
Tangibility		−0.0258(−0.345)	11.6901(0.842)	−0.0054(−0.066)	10.9246(0.744)
Competing offerings	−0.0006**(−2.660)
Inverse Mills ratio		−0.1910(−0.042)	490.9394(0.712)	−1.4545(−0.340)	358.9127(0.560)
Constant	1.4304***(10.208)	10.8599***(10.336)	−36.5431(−0.162)	11.1358***(10.692)	−0.6810(−0.003)
Observations	847	433	357	433	357
R 2	0.4145	0.3860	0.1963	0.3931	0.1773
Year FE	N	Y	Y	Y	Y
Industry FE	Y	Y	Y	Y	Y
Platform FE	Y	Y	Y	Y	Y
Region FE	Y	Y	Y	Y	Y

Note. This table controls for the potential selection bias stemming from campaign success. Column (1) includes coefficients from the first-stage Heckman selection model, where success dummy is the dependent variable. Columns (2) to (5) include regression coefficients from second-stage Heckman models. The dependent variable in Columns (2) and (4) is the natural logarithm of total amount raised [ln(Amount)] and in Columns (3) and (5) is employee growth. Heteroscedasticity robust t-stats clustered at the region level are reported in parentheses. All variables are defined in Appendix A. EOV = environmentally oriented venture; FE = fixed effects.

*

p ≤ .1. **p < .05. ***p < .01.

We further apply the coefficient stability test by Oster (2019) to the main models. We apply it to models (2) and (4) of Tables 2, 3 and 5, respectively, to test for the severity of potential omitted variables. Following the recommendation of Oster (2019), we use R_Max^2 = 1.3 × R^2 as the maximum R^2 achievable in the models to create a lower bound of beta. R^2 is the R² from the controlled models estimated above. We use δ = 1 when creating a lower bound for beta with the assumption that the observables are at least as important as the unobservables. Tests in Supplemental Table IA12 show that the corrected betas remain positive and with similar signs as in the original tests. This indicates that observable variables are at least as important as unobservable variables, suggesting that omitted variables do not play a dominant role in driving the main results. We also apply the more restrictive test by Masten and Poirier (2026), to assess the lowest level of unobservable selection (r_x) to observables that would be needed to change the sign of the estimate. The highest attainable level of r_x is 1, our r_x estimates range from 0.58 to 0.82, implying that our results are moderately to strongly robust to potential omitted variable bias.

Next, we test if the EOV—Employee growth effect persists after using alternative measures of employee growth. We include the difference in employees between t and t + 1, and the percentage growth in employees from t to t + 2. Our findings in Supplemental Table IA13 do not alter our previous conclusions. In addition to accounting for the actual headcount growth, we also acknowledge that using a percentage measure leads to large percentage increases in employment for a small change in the actual number of employees. To further ensure that this does not drive our findings, we split the sample between large and small firms in Supplemental Table IA14. The effect persists in three out of four specifications.

Next, we test if we observe an effect of EOVs on investments in other assets (working capital and physical assets) as well. Investments in working capital and physical assets should be less dependent on the signaling value of environmental orientation. Therefore, we would expect a zero or negative effect on such investments as firms redirect resources toward employee growth rather than investments in working capital and physical capital. Our tests in Supplemental Table IA15 do not show any relationship between EOV and Asset growth.

Lastly, the use of green labels may lead to misclassification; to partly address this issue, we create a third measure of environmental orientation: EOV common. EOV common includes the firms that are classified by both the Beauhurst classification and our text-based measure. We argue that double classified firms with greater certainty are true EOVs, as they have both passed Beauhurst’s manual screening and include environmentally oriented keywords in their company descriptions. ¹² Our findings in Supplemental Table IA16 do not alter our previous interpretations; EOVs raise more capital and hire more people in the post-campaign stages.

Policy Implications

Our findings have implications for policymakers, platforms, and entrepreneurs. Our findings are insightful considering that small businesses account for the majority of employment in the economy (61% of total employment in the United Kingdom according to the Federation of Small Business report [2023], and 80% of the net job creation in the United States [Bowers, 2024; Press, 2024]). Ventures are important for employment but at the same time often face difficulties in obtaining financing to realize their intended hiring. We show that ECF as an alternative source of capital for young ventures is associated with subsequent employee growth. As job creation is viewed as one of the main governmental goals, our findings highlight the need to facilitate the flow of capital from investors through ECF platforms to ventures. Policymakers play a key role in promoting trust in otherwise opaque ECF investments through legal harmonization and standardization across platforms. In countries lacking venture capital, business angels, and bank financing for young ventures, governments could provide tax benefits for ECF investments (Cicchiello et al., 2019), and especially for EOVs to both promote job creation and environmental sustainability.

For ECF platforms, our findings underline the potential of environmental labeling in connection to post-campaign outcome reporting as a mechanism to enhance investor engagement and trust. Platforms highlighting environmental orientation and at the same time showing post-campaign employment outcomes could attract a broader base of impact-oriented investors, which should improve liquidity, campaign success, and in the end platform growth and profitability (Cumming et al., 2025).

For entrepreneurs, our results show that environmental orientation is not only a personal preference but can also be an important early-stage organizational capability that is valuable in capital and labor markets. Highlighting the venture’s environmental orientation in campaign communications and job ads may work as a credible signal to both investors and prospective employees, which facilitates attracting both financial and human capital in the absence of tangible assets or a track record.

Limitations

Our study is not without limitations. The first concerns the small number of EOVs. The Beauhurst measure only categorizes 7.10% of the firms as EOVs or 47, while the second measure categorizes 13.20% or 87 as EOVs. Even though we winsorize the data, the sample size is limited and become smaller with the inclusion of accounting controls, making our outcomes sensitive to outliers. This is also evident when we study the absolute, relative, and total employee growth numbers. Where the economic magnitude in percentage growth ranges from 47.78% to 61.95% for our first measure and 18.06% to 28.58% for our second measure, the effect only ranges from 0.94 to 3.25 employees in absolute terms. Thus, the incremental impact of EOVs on job creation is limited. Even when aggregating the EOVs in our sample, the total job creation ranges from 77 to 168 jobs the year after or roughly 29 to 42 jobs per year created from EOVs relative to NEOVs. This further highlights not only the limited size of the ECF market in its current form but also the potential impact it may have on employment if it continues to grow in importance.

We further acknowledge that our study is conducted in a well-developed ECF system. The ECF system in the United Kingdom is concentrated in two main platforms capturing 90% of the market according to CMA UK 2021. ¹³ Hence, there is less room for specialized exchanges focusing on EOVs, where the platform already has done parts of the due diligence on environmental orientation. Furthermore, we strictly focus on ECF offerings, which are larger and should have a greater impact on economy-wide employment relative to reward-based crowdfunding that may mainly foster self-employment.

It is important to note that our mechanism tests should be viewed as suggestive rather than conclusive and causal. The evidence in support of a supply-side mechanism is not strong and should therefore be interpreted with caution. There are likely other mechanisms at play that we do not capture with our supply- and demand-side proxies. Hence, we can only argue that our findings suggest the presence of supply-side factors impacting employment in EOVs.