This is our eighth blog post for our Job Market Paper Series blog for 2024-2025.
Florian Anselm Munch is a postdoc at the London School of Economics. His research focuses on technology adoption and innovation, and export and upgrading in value chains. You can find his JMP here.
Export-led growth has been the main developmental path advocated for the last 40 years. Export allows firms in low- and middle-income countries to sell to richer clients at higher prices, thereby improving productivity, quality, and management (Atkin et al. 2017, Alfaro-Urena et al. 2022, Bloom et al. 2021, Goldberg and Reed, 2023, Verhoogen 2023). But how do firms become exporters in the first place?
Export markets are dominated by a few large companies (WDR, 2020; Freund and Pierola, 2015; Zavala, 2022). Export requires an upfront sunk-cost investment, e.g., searching for clients abroad and understanding regulations and international logistics. Few small firms manage to overcome these costs and export. Female entrepreneurs incur a double burden: they manage small, less productive companies (Alison et al. 2023, Fang et al. 2022) and face additional barriers (Ackah et al. 2020, Ubfal 2024). For example, societal expectations for care work compromise their ability to network and travel abroad, and low numbers of female entrepreneurs mean few role models and the necessity to prove their entrepreneurial ability to third parties, e.g., bureaucrats or loan officers.
This raises the question of how small firms can benefit more from export and globalization opportunities and how to address the additional barriers for female entrepreneurs?
In my job market paper, co-authored with Amira Bouziri and Fabian Scheifele, we test whether small female-led firms could overcome the barriers by forming a female-only consortium. Partnering with the Tunisian Export Promotion Agency and the German Corporation for International Cooperation, we conduct a randomized controlled trial with 176 female entrepreneurs in Tunisia. The women manage or own a formally registered company with a median of 7 employees. 60% have never exported but are interested, and 40% have.
We randomize the 176 female entrepreneurs into a pure control group of 89 firms and a treatment group of 87 firms. Among the 87 firms in treatment, we form four groups along sectoral lines: Digital Services, Consulting, Agro-Food, and Handicrafts and Cosmetics. Within each sector/consortium, the firms produce different, complementary products, which implies they are not direct competitors. They are located across Tunisia, with a concentration in the country’s three urban economic centres: Tunis, Sfax, and Sousse.
How did we incentivize entrepreneurs with no prior relationship to cooperate and establish a consortium?
In the first year, solely the female entrepreneurs in the treatment group were invited to a series of 8 half-day group-level workshops run by a team of local consultants, all but one female. The workshops were organized by separately for each of the four sub-groups and were designed to create trust, provide export and management knowledge and networking opportunities, feature inspirational talks from successful female entrepreneurs, and outline options for interfirm cooperation. At the end of this period, all four groups decided to create a consortium.
In the second year, the treated female entrepreneurs who decided to sign the contract creating the consortium as a legal corporate group take the lead and determine the consortium’s activity. The implementing agencies do not organize regular meetings anymore but provide ach consortium with a subsidy of 75,000 Euros to kickstart its activities and cover part of the fixed costs of export and coordination between the firms. The subsidy is only announced after the signature of the legal contract creating the consortium to reduce moral hazards. It is earmarked for specific expenditures, e.g. recruiting an accountant and coordinator, developing an online appearance, and sending 2-3 consortium members to 4-6 trade fairs or B2B missions.
What are the effects of the intervention on firms’ exports and business performance?
Firms in the treatment group are 16.5 percentage points (pp) [CI: 3.5-29pp] more likely to have exported in 2023 – 2024 in July 2024, the second period of the intervention (Figure 2). This corresponds to a 59% increase relative to the control group in the same period. These effects are large. For example, business intelligence increases Danish firm’s export likelihood by 8 pp (Munch & Schaur 2018), trade fair visits Japanese firms’ export likelihood by 8 pp [CI:3.9-12pp] (Makioka 2021), while information in Vietnam (Kim et al. 2018) and management consulting (Iacovone et al. 2023) did not affect firms’ export likelihood.
The consortia increased treated firms’ sales by 146% on average in 2023, the second year of the intervention. For example, the median firm in the treatment group has a growth rate of 100% of total sales relative to the baseline, while the same number for the median control firm is 40%. On aggregate, treated firms report an additional 5 million Tunisian dinars, about EUR1.5 million, resulting in roughly 1 million Tunisian dinars (EUR300,000) in additional tax revenue from value-added tax, implying the program paid for itself within two years.
What are the key advantages of the consortium as opposed to supporting individual entrepreneurs?
First, there is cost-sharing. The female entrepreneurs collectivize the costs of searching for new business and export opportunities and jointly market their products. One to three members are sent to represent the consortium who share business cards and opportunities with the other members via smartphone, allowing the entrepreneurs to review more markets quicker. This approach is also cost-effective for the implementing agency, reducing its costs by two-thirds.
Second, the consortia enlarge female entrepreneurs’ networks, inducing them to share human and social capital, and encouraging each other. Consortia members regularly discuss business challenges with an additional six people, three of whom are other female entrepreneurs whom they met via the intervention. Female entrepreneurs with smaller pre-intervention networks – a few had no contact with any female entrepreneurs – benefit particularly. Figure 3 illustrates consortia members are 39, 36, and 47 pp more likely to share management practices, product ideas, and export experience respectively, with other entrepreneurs, and 32 and 46pp more likely to bid with other entrepreneurs for a large contract and receive emotional support vis-a-vis the risks of business and export from other entrepreneurs. Moreover, female entrepreneurs consistently express a heightened belief in their entrepreneurial abilities.
Two considerations suggest room for improvement in similar interventions in the future. First, cooperation has its challenges and is not for everyone. Only 45% of the invited firms remain in the consortia after two years, about 8-12 firms per group. While the first period was dominated by an improved perception of partnership and cooperation between female entrepreneurs, this effect vanished until the endline, driven by conflicts over joint decision-making and effort. Second, peer quality and homophily matter. With a low initial knowledge level, e.g., of exporting, external knowledge inputs from consultants into the network are essential. Firm heterogeneity becomes a problem when incentivizing firms to do business together; female entrepreneurs are interested in cooperating with firms that are useful for their business.
What are the big-picture implications and limitations?
Small firms in low-and middle-income countries can be incentivized to and benefit from sharing the fixed costs of export. This might apply to other fixed costs too, such as for technology adoption, take other forms of organizing collective action, such as contracts or cooperatives. This is seemingly a promising avenue to generate economies of scale while preventing small firms dying. While such contracts or organizations could and did historically occur via interaction between private agents, coordination failures, information asymmetries, and low levels of trust provide a rationale for government intervention to coordinate collective action between firms.
Our study has the following limitations. The sample size is small, calling for caution and replication. The effects on networks, confidence, and management practices might be context-specific, potentially limiting external validity. Given the bundled nature of the treatment, we can’t disentangle each component’s contribution to the effect. For example, did female entrepreneurs’ belief in their entrepreneurial abilities increase due to the knowledge transfer from the consultants, support from the government agencies, or the exchange with the other consortia members? We think combining all inputs was necessary, but we can’t rule out that it is otherwise.
Feature Image Source: Implementing agency – GIZ
