Policy Research Working Paper 9552 Business Training and Mentoring Experimental Evidence from Women-Owned Microenterprises in Ethiopia M. Mehrab Bakhtiar Gautam Bastian Markus Goldstein Africa Region Gender Innovation Lab February 2021 Policy Research Working Paper 9552 Abstract Recent research shows that microenterprises in develop- business training received customized mentoring from ing countries are constrained by their managerial capacity, these “trained mentors.” Pooled results using three rounds especially in the areas of marketing, record keeping, finan- of post-training surveys carried out over three years show cial planning, and stock control. In a stratified randomized that business training causes profit and sales to improve by controlled trial, experienced businesswomen in Ethiopia 0.21 standard deviation, while business practices improve were given a formal business training that addressed these by 0.13 standard deviation. The overall impact of mentor- constraints. A second-stage mentoring component in which ing is muted—strong impacts are observed on the adoption a random selection of female mentees within the social of business practices among mentees, but there is no statis- and business network of the trainees from the first-stage tically significant impact on profits. This paper is a product of the Office of the Gender Innovation Lab. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/prwp. The authors may be contacted at mgoldstein@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Business Training and Mentoring: Experimental Evidence from Women-Owned Microenterprises in Ethiopia M. Mehrab Bakhtiar∗ Gautam Bastian IFPRI Independent Researcher Markus Goldstein World Bank JEL: D22, L26, M53, C93, J24 Keywords: Entrepreneurship, Business training, Mentorship, Impact Evaluation, Field Experiments, Gender ∗ Contact: M.Bakhtiar@cgiar.org; gautamb@gmail.com; mgoldstein@worldbank.org. We thank the Umbrella Facility for Gender Equality for financial support and seminar participants at the Pacific Conference for Development Economics (PACDEV), 2017; Centre for the Study of African Economies (CSAE) Conference, 2018; Annual Bank Conference on Africa (ABCA), Stanford Uni- versity, 2018 for their helpful comments. Eliana Carranza was closely involved with the Impact Evaluation design and provided insightful feedback throughout the duration of the research project. Adiam Hailemichael, Nikita Arora, Tigist Ketema, Ombeline Gras and Brittany Hill provided ex- cellent research assistance as well as field and administrative support. Vanessa Adams and Hebret Abahoy - from ACDI/VOCA, the implementing partner in this project - were superbly collaborative. We are indebted to David Mckenzie, Salman Alibahi, Joao Montalvao, Kenneth Leonard, Rachael Pierotti and Michael O’Sullivan for their invaluable feedback. The views presented in this paper are the authors’ own and do not represent the World Bank and its member countries. 1 Introduction Ownership of, and employment in, microenterprises account for a large fraction of the labor market in the developing world. In many countries, microenterprises underperform other forms of businesses in terms of productivity and profitability. One possible explanation for the dismal outcomes of microenterprises is that microenterprise owners lack managerial capital, which is the skill-set required to run a business (Bloom and Van Reenen, 2007; Bruhn, Karlan, and Schoar, 2010; McKenzie and Woodruff, 2016). Policymakers have attempted to address this constraint that small businesses face by scaling up business training programs all over the developing world. While an emerging body of literature finds that business training can improve management practices, the measured impact of standard versions of this training on business profit or operational scale appears to be modest, at best (McKenzie and Woodruff, 2016; Quinn and Woodruff, 2019). McKenzie and Woodruff (2016), however, argue that the relationship between ‘better’ man- agement/business practices1 - particularly, in the areas of marketing, record keeping, financial planning and stock control - and performance is causal and that the lack of an observable effect in the impact evaluations of most business training programs is because firms do not change business practices enough rather than because these practices do not matter.2 More- over, in a critical review of the literature, McKenzie and Woodruff (2014) find that many of the business training evaluations measure impacts only within a year of training. They argue that such short-term evaluations are problematic because it may take longer than a year for adopted business practices to translate into an impact on business outcomes. Thus, several recent studies had an extended period of follow-up after training interventions were completed (Bruhn, Karlan, and Schoar, 2018; Higuchi, Mhede, and Sonobe, 2019; Karlan, Knight, and Udry, 2015; David McKenzie Susana Puerto, 2017; Valdivia, 2015). However, these studies show mixed impacts of business training programs on business performance. 1 The study uses 26 questions that measure business practices in the areas of marketing (e.g., Does the firm advertise? Does it attempt to attract customers with a special offer? Does it ask customers what other products they would like it to sell?), record keeping (e.g. Does the firm record every sale and purchase? Has it worked out the cost of each item it sells? Does it have a written budget?), financial planning (e.g. Does it have a sales target for the next year? Does it have a balance sheet and profit and loss statement?), and buying and stock control (e.g. Does it frequently run out of stock? Does it attempt to negotiate discounts with suppliers?). These are intended to be universal best practices, in the sense that most firms should benefit from using them. They are closely based on the goals of business training programs like the ILO’s Improve Your Business (IYB) program. 2 Importantly, McKenzie and Woodruff (2016) show that the (modest) effects observed in several business training programs on profits/sales are remarkably consistent with the predicted effects given the observed changes in business practices. 1 To address this gap in the literature, we evaluate the impact of a well-designed and intensive business training curriculum on business practices and performance over an extended period. The business training program targeted high-potential women entrepreneurs in five regions of Ethiopia and was carried out as a randomized controlled trial. Pooled results using 3 rounds of post-training surveys, carried out over 3 years show that business training causes profits and sales to improve significantly. This effect appears to be driven by the adoption of several of the ‘better’ business practices identified by McKenzie and Woodruff (2016) that were covered extensively in the business training program. Given the extent of micro-entrepreneurial activities in developing countries, it is, however, important to identify low-cost programs that can lead towards the adoption of these business practices with the aim of subsequently improving business performance. Standard business training programs, like the one mentioned above, are expensive - business training programs can cost hundreds of dollars per student and collectively, several billion dollars have been spent behind business training all over the world involving millions of trainees (Blattman and Ralston, 2015).3 This problem could be potentially addressed by a mentoring intervention involving trained mentors: once more experienced business people are given formal business training, they can be connected with smaller firm owners from their networks to pass on what they had learned in the training program as well as provide customized mentorship. This is a promising low-cost approach to business training programs that organically disseminates relevant market information and advice on business practices and knowledge to new and fledgling businesses. To understand the impact of mentoring, business women (mentors) who completed the busi- ness training were assigned to provide mentoring to a random subset of smaller firm owners (mentees) within their social and business networks that they had nominated when applying for the business training. We find that mentees, randomly assigned to receive mentoring, exhibit strong effects on business practices. Profits, however, were not significantly affected. However, we show that, despite lacking statistical significance, the effect of mentorship on profit is remarkably consistent with the predicted effect given the observed changes in busi- ness practices. This study is similar to the impact evaluation of a training and a mentoring intervention by Brooks, Donovan, and Johnson (2018), where they directly compare the impact of one- to-one mentorship to a traditional business training. They find that in the short term, one-to-one mentoring results in higher business profits as a result of a transmission of local 3 For example, the International Labor Organization’s Start and Improve Your Business program has trained more than 4 million people in 100 countries since 1977. 2 market information, such as low-cost suppliers. However, the effect appears to fade in the longer term. Our study departs from Brooks, Donovan, and Johnson (2018) in two important ways. First, we evaluate whether firms that receive formal business training, adopt ‘better’ businesses practices. Here we focus on marketing, stock-keeping, record- keeping, and financial planning business practices that McKenzie and Woodruff (2016) have identified as essential inputs for improving business performance. We specifically examine whether changed practices translate into better outcomes for the firms especially in the longer term (i.e., approximately three years after the training). Secondly, if business performance is indeed constrained by the adoption of these practices, receiving mentoring alone - as in Brooks, Donovan, and Johnson (2018) - may not induce improved business performance in the medium and long term. However, such an effect maybe possible through the provision of mentoring by trained mentors – provided that communication of the better business practices happens effectively through the mentoring. Together, the formal business training program and the mentoring component is known as the Ethiopia Women Agribusiness Leaders Network (WALN) project. The evaluation of this project allows us to investigate the importance of business practices such as marketing, stock- keeping, record-keeping, and financial planning for small, women-owned firms in Ethiopia, and the longer-term consequences of adopting these practices on business performance. The remainder of this section provides a more comprehensive review of the literature. Section 2 describes the WALN project. Section 3 lays out our experimental design, with section 4 explaining our empirical approach. Section 5 presents our results and section 6 concludes. 1.1 Additional Literature Review Business and Management Training Business training is the broader approach encompassing business skills training, mentoring and consulting interventions to improve the poor business skills and management practices that are common in developing countries. These inefficiencies more severely affect women who constitute a larger share of self-employment in small enterprises and often see lower returns than their male counterparts (Calderon, Cunha, and Giorgi, 2020; Gine, Xavier and Mansuri, 2014; De Mel, McKenzie, and Woodruff, 2012). A number of impact evaluations of business training programs have yielded mixed results. A study in rural Mexico increased the use of formal accounting techniques, and improved 3 revenue, profits and the number of clients of female business owners (Calderon, Cunha, and Giorgi, 2020). In urban Sri Lanka, on the other hand, although training improved the business practices of self-employed women, there was no effect on business profit, sales or capital stock. When the business training was coupled with a grant however, there was a significant—albeit short-term—improvement in business profitability (De Mel, McKenzie, and Woodruff, 2012). Combining business training with large loans (about seven times the average loan size) to microfinance clients in rural Pakistan, Gine, Xavier and Mansuri (2014) find that although business training alone increases business knowledge, improves business practices (such as record-keeping of sales), and increases household expenditures, group cohesion and general outlook on life these beneficial effects are largely reaped by male clients. Women clients do show a significant improvement in business knowledge, however they have little control over their businesses— 40 percent reported that their (male) spouses are responsible for most of their business decisions. Adding the loan to the training has little effect irrespective of client gender, perhaps indicating that existing loan sizes already meet the demand for credit. Karlan and Valdivia (2010) find that providing 30 to 60 minute business training to female microfinance clients in Peru improved business knowledge and client retention for the MFI, but there was no impact on business outcomes such as profits or revenue. Valdivia (2015), studying another Peruvian sample, finds that personal development, business management and productive skills training intervention aimed at empowering women and improving their access to credit and their business practices has some interesting effects. Women who were offered an additional three-month technical assistance package were more likely to make business process innovations and saw an 18% increase in sales. By comparison, women who did not receive technical assistance were more likely to close their businesses. McKenzie and Woodruff (2016), mentioned earlier, demonstrate that business training pro- grams which have larger effects on management practices also have larger impacts on profits. They suggest that most entrepreneurship training programs appear to fail because they do not improve business practices sufficiently, not because business practices themselves do not matter. Some more recent work on training has emphasized the importance of curriculum choices. For example, Campos et al. (2017) compare a standard business training and one that emphasizes personal initiative. The personal initiative training focuses more on mindset than particular practices. They find that the personal initiative training shows significantly higher profits than standard training, and that this difference is more pronounced for women. 4 In another innovation in the design of curriculum for business/management training, Bloom et al. (2013) demonstrate that individualized consulting programs can be effective. Using high-frequency data from large textile factories in Mumbai, India, Bloom et al. show that an intensive consulting intervention greatly improved output per worker and quality. They also show that management practices (specific to textile production) improved significantly. Finally, some emerging evidence indicates that individual business consulting may improve managerial skills and self-confidence of owners of small and medium enterprises (Bruhn, Karlan, and Schoar, 2018 in Mexico) and business literacy (Karlan, Knight, and Udry, 2015 in Ghana). A randomized controlled trial in the Dominican Republic has also shown that simpler business advice is more efficient in improving outcomes for lower-profile participants (Drexler, Fischer, and Schoar, 2014). In summary, existing evidence suggests that the impact of business training and consulting is highly sensitive to geographical and business contexts as well as to curriculum content and add-on interventions such as technical assistance and finance. In light of this evidence, it is important to understand the specific business environment, needs and constraints faced by local entrepreneurs before implementing a business skill training program. Special attention needs to be paid to the specific needs of women entrepreneurs, especially their need for training and finance. The importance of one-to-one counseling as well as the need to tailor advice to the level of the participants should be kept in mind; making WALN’s mentoring component a particularly relevant alternative to classical business training. Mentoring The appeal of mentoring, as an intervention oriented to bridging gender gaps, lies in its ability to fit the needs of women-owned and women-managed small and medium enterprises, for whom a lack of positive gender-role models may be a significant barrier to participating in high-earning male-dominated sectors and to overcoming productivity and earnings gaps even when they are able to participate (Campos et al., 2015). In this section we briefly situate the WALN mentoring component in the theoretical and em- pirical literature on mentoring in general and in how it relates to women’s entrepreneurship. The current conceptual framework about mentoring is based on Kram (1988; 1983) who views mentoring as a combination of career oriented advice and behavior modeling and personal psychosocial support that facilitates goal oriented knowledge acquisition, social learning and behavior change. The mentor serves as a role model for effective business behavior (Allen et al., 2004) while the mentee receives career and psychosocial (self-esteem) benefits. 5 A Canadian study by St-Jean and Audet (2012) finds that mentee entrepreneurs benefit most significantly on cognitive and skill-based learning, e.g., management knowledge and skills, improved vision for their business venture and identifying new opportunities. Mentees also experience affective learning benefits to a lesser degree, such as an improved self-efficacy, validation of entrepreneurial self-image and a lowered sense of solitude, factors that could in turn influence entrepreneurial resilience. St-Jean (2012), using another Canadian en- trepreneur mentoring dataset, finds that the career-related functions of mentoring are the most effective factor in the development of learning, followed by psychological and role-model functions. Trust, perceived similarity and mentee self-disclosure are found to be essential to build a highly effective mentoring relationship. Several recent studies have experimentally tested the impact of effective mentoring (or mentoring-like) which are cost effective (with some being better suited for less-experienced businesses). As referred to earlier, Brooks, Donovan, and Johnson (2018) find that in the short term, one-to-one mentoring results in higher business profits for less experienced mi- croenterprises. In this non-incentivized setting, mentees who continue to meet with their mentors 12 months after the official end of the treatment period are more profitable than those who are no longer meeting. Despite a clear selection issue, this finding highlights the possibility that introducing incentives in mentoring interventions may increase the intensity and durability of the mentoring interaction, which, in turn, can improve longer-term business performance. Lafortune, Riutort, and Tessada (2018) compare the impact of a personalized consulting intervention and a role model-type intervention for microenterprises in Chile. Both programs have large impacts (which are similar in magnitude) on business profits, with the latter shown to be more cost effective. Another recent study on mentoring is a long-running field trial involving junior female economists in the USA which experimentally varied access to mentoring workshops on advancing research careers in academia (Ginther et al., 2020). Treated women are found to be more likely to stay in academia and more likely to have received tenure in an institution ranked in the top 50 in economics in the world. The theoretical literature on social networks suggests that beliefs and ideas are more likely to be changed when the changes originate from trusted social bonds (Jackson, 2010). In traditional societies, homogeneous networks may reinforce social norms, in turn reinforcing current behavior and attitudes, especially for women. Diversifying networks may help to connect women to new ideas, role models and business contacts, even encouraging greater innovation. Kandpal, Baylis, and Arends-Kuenning (2012) find that a women’s empower- ment program intervention in northern India was able to affect social norms by increasing 6 participants’ social contacts outside their own caste group. They also find that participating women have higher physical mobility, political participation, and access to employment as well as spillover effects on non-participants relative to women in untreated areas. Conley and Udry (2010), Bandiera and Rasul (2006), Munshi (2004) and Foster and Rosen- zweig (1995) measure the consequences of social learning in various contexts. More recently, BenYishay and Mobarak (2019) and Beaman et al. (2018) experimentally vary the introduc- tion of information in social network nodes and measure its impact in information dissemina- tion and adoption. Related to our study, Atkin, Khandelwal, and Osman (2017) and Cai and Szeidl (2018) measure the impact of potential knowledge transfers from exogenously selected partners among (relatively larger) businesses. Atkin, Khandelwal, and Osman (2017) study the diffusion of knowledge in the context of supplier-customer relationships by randomly allocating new foreign rug orders to Egyptian rug makers, and Cai and Szeidl (2018) study the creation of randomly formed business groups in China. Both studies find a sustained impact on profits as a result of the interventions. The goodness of fit between the mentor and mentee has been identified in the literature as a key ingredient of a successful mentoring intervention. The WALN project leverages existing social ties between mentors and mentees, i.e. mentors identify promising mentees, helping overcome lack of trust which is a typical barrier to information dissemination. Sim- ilar to Brooks, Donovan, and Johnson (2018), the WALN mentoring component was not incentivized. Also similar to Brooks, Donovan, and Johnson (2018), in the WALN project, both the mentors and mentees are women, which may provide a beneficial support structure. Es- pecially in the male-dominated Ethiopian agribusiness sectors, women-to-women mentoring may help reinforce achievement-oriented behavior, providing task-specific feedback and al- leviating stress (Noe, 1988; Burke and McKeen, 1990; Nelson and Quick, 1985). Campos et al. (2015) hint at a counter-argument that men with existing networks in male-dominated sectors might be able to open more doors for women mentees, but their argument is more about business entry rather than success conditional on entry. Overall, there is no consensus in the literature on the relative effectiveness of women and men mentors for women mentees (Hansford, Tennent, and Ehrich, 2002). In summary, there is a broad acceptance in the literature of the idea that mentoring is a useful and effective method of transferring knowledge, and improving business and career outcomes, especially since each program can be adapted to fit the needs of the participants. 7 2 Program Description The Women in Agribusiness Leaders Network (WALN) project involved a first-stage of busi- ness training and a second-stage of mentoring—which was carried out by the business women trained in the first-stage. The randomized controlled trial design of the program led half of the potential mentors and mentees—who were eligible to participate in the WALN project— to be randomly assigned to receive business training and mentoring interventions, respec- tively. The other halves formed the control groups that did not receive any intervention. Comparing the treated groups to the control groups allows us to separately measure the impact of the business training program and the mentoring program. 2.1 Program Overview WALN was embedded in the Agribusiness and Market Development (AMDe) component of the Government of Ethiopia’s USAID and World Bank-funded Agricultural Growth Program (AGP) and was part of USAID’s wider Feed the Future programming. WALN aimed to improve business skills of female participants, enabling them to be community leaders and change makers. The particular focus of the WALN project on women was guided by USAID/Ethiopia’s mission-wide gender analysis which found that while women constituted 45% of the agricul- tural labor force, they accounted for only 20% of the members of agricultural cooperatives and had less access to productive resources and opportunities than their male counterparts. Similarly, Aguilar et al. (2013) analyze household survey data from rural Ethiopia and find a 23% gender gap in productivity (in terms of gross value of output) per hectare between female farm managers and male farm managers. WALN sought to improve agribusiness outcomes by addressing gender differences in produc- tivity, profitability, participation and leadership in the sector. The intervention was based on the idea that training, mentoring and networking could expand women’s capacity to play leadership roles in sectoral organizations and manage business profitably. WALN was ex- pected to improve women’s participation in leadership and decision-making in the overall agriculture sector in Ethiopia. A focused business and leadership training, as well as a mentoring program may help women strengthen and expand a fledgling support network to expand economic empowerment. In- 8 creased female membership and leadership in cooperatives, more savings and better market- ing may result as more women are visible, credible and effective business leaders. Bringing the voices of women to the table more effectively impacts issues such as health, nutrition, early marriage and household financial management. The Africa Gender Innovation Lab of the World Bank partnered with USAID and the Gov- ernment of Ethiopia to evaluate the impact of WALN. The impact evaluation of WALN was designed to capture the business and social-network effects of business training and mentoring on mentors and mentees. 2.2 Description of the Interventions The WALN project was composed of three components: (i) a business training course; (ii) a mentoring system; (iii) establishing a formal network of women entrepreneurs. The objective of this project was to engage high-potential women business leaders to achieve the following objectives: (1) build skills in key areas such as negotiation, marketing, net- working, financial planning and communication; (2) develop enhanced leadership capabilities that will enable them to grow and manage their business more profitably, (3) develop a pro- fessional network with women leaders in agriculture and related business, and (4) serve as mentors to aspiring businesswomen in their networks or community and be a role model. WALN focused on 9 specific value chains that were part of AGP-1, namely chickpeas, coffee, honey, maize, sesame, wheat, livestock, dairy and teff. Businesswomen in these value chains working in 96 targeted woredas across 5 regions in Tigray, Amhara, Oromia, SNNPR, and Addis Ababa were eligible to participate. Business and Leadership Training This training package sought to provide basic business management skills that are essential for running profitable and sustainable businesses. Intensive business and leadership training was provided, to test if it could expand women’s capacity to play leadership roles in sectoral organizations and manage businesses more profitably. International and national business specialists were called upon to deliver this training using adult-learning methods such as case studies, facilitated discussions, group work, interactive presentations, plenary sessions, role- play, individual exercises and games. Between July and December 2014, 99 women received approximately 60 hours of classroom training spread over 6 sessions of 2 days each on the 9 four skill areas/modules detailed in Appendix Table A.3. These women were referred to as ‘WALN Mentors’. All modules except Module IV that focuses on Mentoring and Coaching were meant to improve the mentor’s business. Module IV focused on giving mentors the skills to advise mentees. Further details on the business training curriculum are available in Appendix Section A.3. The training curriculum was developed by technical experts hired by ACDI/VOCA. It is a fairly standard curriculum for these types of training, but it does not explicitly follow some of the more established curricula like EMPRETEC or ILO SYB.4 The focus on mentoring and the integration of a mentoring component in the program are the more innovative and novel elements of this intervention. Compliance to the business training program was quite high. Approximately 94% of the businesses randomized to the business training/‘WALN Mentors’ treatment report to have attended at least one training session. Moreover, average, treatment mentors participated in 5.2 sessions (out of 6 sessions in total). Mentors, who completed the business training, reported mentoring 1.3 mentees on average. Only 10% of these mentors report that they did not provide mentoring to anyone after completing the business training. Mentoring Program Generally, mentoring is a professional relationship in which a mentor, an experienced person, assists a mentee, a less experienced person seeking help and advice, to develop specific skills and knowledge that will enhance the latter’s professional and personal growth. In the setting of WALN, the mentoring component leveraged existing social ties between the mentors and mentees to deliver elements of customized business training and counseling in a semi-formal setting. During the application for the WALN business and leadership training programs (later to become potential mentors) applicants were asked to nominate potential mentees from among their social networks. On average each mentor nominated close to 7 mentees. Within each 4 EMPRETEC is a capacity building program by the United Nations Conference on Trade and Devel- opment (UNCTAD) to ‘promote the creation of sustainable, innovative, and internationally competitive small- and medium-sized enterprises (SMEs).’ The Start and Improve Your Business (SIYB) program is a management-training program developed by the International Labour Organization (ILO) with a focus on starting and improving small businesses as a strategy for creating more and better employment for women and men, particularly in emerging economies. With an estimated outreach in over 100 countries, it is one of the world’s largest program in this field. 10 mentor’s nominees, on average 3 mentees were randomly assigned to receive mentoring. Mentees were typically younger women, who wanted to start their own agribusiness, were in the informal sector, or those who were not yet competitive in the agribusiness market but had the potential. The intervention encouraged these women to formalize their businesses and become better business managers. Once mentors had undergone the business and leadership training described above, they were asked to use their knowledge to mentor and coach their prospective mentees who were randomly assigned to receive mentoring for 6 months. The project used a series of intensive, in-the-field workshops and mentor-mentee sessions, along with ongoing coaching and sup- portive contact among women in their network. Mentors received ongoing oversight from the project staff, including site visits to observe progress and provide targeted technical as- sistance – leveraging, where possible, existing project efforts and staff. Time was set aside at the quarterly network meetings for mentors to share lessons learned and challenges faced in order to improve their performance. Compliance to the mentorship program is moderately high. Of the businesses randomized to receive the mentoring treatment, 77.6% report to have attended at least one session with their mentors. Business Convention/Networking The third component of this intervention created networks and linkages between mentors and mentees at sessions, forums and workshops, including encouraging the use of digital networking via email, blogs and social network. This created opportunities to develop new businesses, learn about new opportunities, solve problems and increase visibility for both, mentors and mentees. As a result, networks were created at the regional as well as national level, and 87.9% of treatment mentors and 68.4% of treatment mentees attended at least one regional networking session. 2.3 WALN Selection Process and Program Cost ACDI/VOCA solicited applications from high-potential women agribusiness entrepreneurs in Ethiopia through advertisements and community mobilization. Women eligible to participate in WALN had to be owners or managers of a registered business or association at the time of enrollment; active at any point in one of the 9 selected value chains in the targeted regions; holders of a bank account (individual or association/business); able to nominate 6-8 11 businesswomen and willing to mentor 3-5 of them; and able to provide a professional letter of recommendation in support of their application. Women who met the eligibility criteria for participating in the WALN project—as outlined in Appendix Table A.2—who completed the application form, provided at least one letter of recommendation and nominated five to eight mentees were considered in the pool of potential mentors who would receive the business and leadership training. Mentees were selected from among the nominated candidates who met the corresponding eligibility criteria. Mentees were assigned only to the mentors who nominated them. To minimize non-compliance and maximize the effectiveness of existing network ties, if multiple mentors nominated the same mentee, the mentee would be allowed to pick the mentor. Altogether, approximately $2,400 was spent per treatment mentor, while approximately $500 was spent per treatment mentee. Because of data limitations, the cost of attending several forums, workshops and conference participation cannot be disaggregated from the basic cost of business training and mentoring. 3 Experimental Design and Data Randomization The pool of eligible applicants became the sample for the baseline survey. Treatment was then randomly assigned to eligible applicants who responded to the baseline survey. The WALN project operated in Ethiopia’s Agricultural Growth Program’s (AGP) target woredas of five regions of Ethiopia: Tigray, Amhara, Oromia, Addis Ababa and SNNPR. The impact evaluation covers the business training and mentoring activities across the entire geography covered by the project. Business training treatment randomization was stratified by region and firm-size tercile. This is the pool that would become the mentors in the second stage, so henceforth we refer to them as such. Half of the mentors received formal business training covering the modules listed above. Treated mentors came to a training facility two days a month for six consecutive months. Control group mentors were not given any business training, but they were tracked and surveyed in each round of follow-up surveys. Half of the mentees of treated mentors were randomly assigned to receive mentoring, while the remaining half were assigned to the control group. After the business training was concluded, 12 mentors were asked to meet their mentees one day a month for six consecutive months. Mentors were specifically encouraged to provide mentoring only to mentees randomly selected to receive mentoring. However, since both treatment and control mentees belong to the mentor’s social and business networks, there was a possibility of spillover across treatment status among the mentees. To address this issue, we include mentees nominated by control mentors in the impact evaluation. This helps us measure the spillover effect of being in a treatment mentor’s network by directly comparing the outcomes of the control mentees (of the treatment mentors) and the pure control mentees (of the control mentors—who did not receive any business training). Appendix Table A.1 shows the sample size for both mentors and mentees across the different survey rounds and treatment status. We report balance tests for the mentor and the mentee samples in Tables 1 and 2, respectively, on a set of individual, household, and business characteristics. The relative size of the mentor and mentee businesses is evident. For example, the monthly (winsorized) profits of mentors at baseline are approximately 3.5 times as large as those of their (potential) mentees.5 Most of the variables are balanced across the different treatments. Imbalance in some of the baseline variables is accounted for by controlling for them in all subsequent (and relevant) regression analyses. Follow-up Surveys Table A.1 as well as Appendix Tables A.4 - A.5 show the WALN impact evaluation design in detail, and in particular, the number of women surveyed in each round of data collection as well as the attrition in the sample from baseline through three post-intervention follow-up data. At baseline, we have a sample of 197 mentors - 99 treatment and 98 control mentors. The mentee sample, on the other hand, consists of 295 treatment, 294 spillover and 539 control mentees. Given the small initial sample of mentors, we placed a premium on follow- up. As observed in Tables A.1, A.4 and A.5, overall attrition is low and does not vary by treatment status. The WALN baseline survey was conducted from March to August 2014. Business training took place from August to December 2014 and this was followed by mentoring sessions from January to July 2015. The first follow up survey took place between August and December 5 The large, albeit statistically insignificant, differences in raw profit and revenue at baseline between treatment and control mentors (as well as between treatment and control mentees) is driven by a handful of (two or three) firms which are mainly large cooperatives. Rerunning all the estimates by dropping these firms produce very similar estimates for all the relevant analysis, implying that these large firms do not drive the main results. 13 2015, while data collection for the second follow-up was carried out in the summer of 2017. Data collection for the final and the third follow up was carried out in February-April, 2018. 4 Outcomes of Interest and Estimating Equations 4.1 Outcomes of Interest Due to the fact that we have a number of variables which capture similar or identical con- structs, we have regrouped the variables of interest into families of outcomes whenever it was possible to do so. For each family of outcomes, we have created an average z-score index by ensuring all variables in the outcome were coded in the same direction, calculating the z-score of each variable by subtracting the control-group mean and dividing by the control group standard deviation, and averaging the z-scores of the outcomes for each family. The individual outcome variables that were used to create the following family of outcomes are explained in Appendix Section A.1. Given the goals of the intervention we focus on a top level of outcome of profits and revenues (combined into an index) and a secondary (mechanisms) index of business practices. To better understand how the intervention worked, the business practice index is broken out into marketing, stock control, record keeping and financial planning indices. 4.2 Estimating Equation: Basic Treatment Effects Business Training We measure the basic (intent-to-treat) program effects of the business training in the follow- ing ANCOVA specification: k yi,t = β0 + βT BusinessTrainingi + β1 Y i,0 + θs 1(i ∈ s) + δj 1(j = t) + εi,t (1) j =1 where yi,t is the outcome variable of interest (e.g., profit) for firm i, survey round, t; θs is the strata fixed effects; δj survey round fixed effects; εi,t is the error term, which is clustered 14 at the firm level (because of firm level randomization within a strata); Yi,0 is the lag of the dependent variable in order to maximize power (McKenzie, 2012). Mentoring We can separately measure the effect of being a treatment and a ‘spillover’ mentee who belong to the social and business network of the same treatment mentor in relation to pure control mentees. This is done to evaluate the spillover across treatment and control mentees of treatment mentors which can bias the treatment effects downwards. We measure the basic (intent-to-treat) effects of receiving mentoring in the following way: k yi,t = β0 +βT TreatmentMenteei +βS SpilloverMenteei +β1 Y i,0 + θS 1(i ∈ S )+ δj 1(j = t)+εi,t j =1 (2) where yi,t is the outcome variable of interest (e.g., profit) for firm i, survey round, t; θs is the strata fixed effects; δj survey round fixed effects; εi,t is the error term, which is clustered at strata level; Yi,0 is the baseline value of the dependent variable in order to maximize power.. The omitted category in the above regression is the (pure) control mentee. 5 Results As indicated in Equations (1) and (2), our primary specification is an intent-to-treat analysis using all the follow-up rounds pooled together, with the baseline value of the dependent variable used on the RHS in order to maximize statistical power. Tables 1 and 2 show that the treatment and control arms in the business training and mentorship programs are balanced at baseline for most of the relevant observables. 5.1 Impact on Mentors We test the hypothesis6 that a business training improves business performance by looking into an index of profit and revenue as well as their individual components (Table 3). Given 6 as pre-specified in our analysis plan 15 the relatively small sample size, we pool impacts over the three post-treatment waves to maximize statistical power, with the coefficients then representing the average impact over 3 years post-treatment. Point estimates suggest large and statistically significant impacts of attending the business training on revenue and profits. The profit and sales index suggests an improvement of 0.21 standard deviation. The estimates for the winsorized revenue and profit variables suggest a pooled impact of 62% and 80%, respectively. Table 4 shows the impact of the business training intervention on business practices that were covered extensively in the training curriculum. Point estimates suggest that treatment mentors fare markedly better in terms of the business practices identified by McKenzie and Woodruff (2016) as well as Bloom and Van Reenen (2007) and Bruhn, Karlan, and Schoar (2010). The business practices index suggests an improvement of 0.13 standard deviation (Table 4), with the largest gain observed for marketing practices (a 0.14 SD improvement over control mentors). Record-keeping and stock control show improvements similar in magnitude to marketing practices while financial planning improves by 0.08 SD for treatment mentors compared to control mentors. Digging deeper into which particular practices may be driving these results, Appendix Ta- bles A.6 - A.9, show the disaggregated results for the underlying 26 business practices for marketing practices, record-keeping, stock control and financial planning. Specifically, treat- ment mentors are 21.4 percentage points (49.5%) more likely to have written business records than control mentors (Table A.8). A similar magnitude of impact holds for keeping finan- cial records for every purchase and sale; using records to calculate cash on hand; and using records to evaluate whether sales of a particular product were increasing or decreasing from one month to another. Treatment mentors were also found to have negotiated or attempted to negotiate with suppliers for a lower price on raw materials -- 16 percentage points (27.4%) higher than the control group (Table A.7). Similarly, marketing practices show strong and positive movements as a result of the business training (Table A.6). Treatment mentors are more likely to visit at least one of their competitors’ businesses to see what prices they are charging or what products they are selling. Treatment mentors are also 17 percentage points more likely to attract customers with a special offer; 19 percentage points more likely to ask suppliers which products are selling well in this business industry; and 20 percentage points more likely to ask at least one former customer to find out why former customers have stopped buying from this business. While a fairly wide range of practices are changing, it is also worth noting that a number of these effects are large in magnitude. In addition to changes in practices, Table 5 documents some evidence of diversification. Treatment mentors have an average of 0.23 more businesses, indicating that the training 16 has spurred some business creation. Qualitative work also indicates that one entrepreneur opened a supermarket as an outlet for the dairy products she was processing. Distribution of treatment effects Figure A.1 shows that, relative to control mentors, the distribution of monthly profit and revenue z-score is shifted to the right for treatment mentors. This suggests that the differ- ences in the means of treatment and control mentors are not being driven by a small number of firms but rather that profits and revenues are generally higher among treatment mentors. A Kolmogorov-Smirnov test rejects the null hypothesis that the two distributions are equal at the 1 percent level. 5.2 Impact on Mentees We test the hypothesis that the mentoring component of this project improves business performance of mentees by looking into an index of profit and revenue as well as their business practices. Since (treatment) mentors were assigned to work with a random subset of their mentees, it is possible that they either spoke about the practices they had learned with all of their potential mentees or that mentees spoke with each other, since they were part of the same pre-existing social network. Given the possibility of spillovers, we compare treatment and spillover mentees within the same regression framework. This allows us to separately measure the effect of being a treatment and a ‘spillover’ mentee who belong to the social and business network of the same treatment mentor in relation to pure control mentees. We also test whether there is jointly no treatment effect and no spillover effect, and whether the spillover effects are equal to the treatment effects. In Table 6, we examine impacts on profits and revenues. Similar to the business training, we pool impacts over the three post-treatment waves to maximize statistical power, with the coefficients then representing the average impact over 3 years post-treatment. Although the point estimates are mostly positive, they are quite small and the impact of the mentoring component is not statistically significant for revenues, profits, or an aggregated index of these measures. In Table 7, we examine the impact of being mentored on business practices. We find statisti- cally significant impacts on almost all of our measured business practices, for both treatment and ‘spillover’ mentees, suggesting strong spillover effects. Point estimates suggest that both treatment and ‘spillover’ mentees do better in terms of some of the key skills identified by 17 Mckenzie and Woodruff (2016) as well as Bloom and Van Reenen (2007) and Bruhn et al. (2010) and which were covered extensively in the first-stage business training (for mentors). Mentors were encouraged to disseminate the importance of the business practices while con- ducting the subsequent mentoring sessions and we see an indication that the knowledge permeated their networks. The business practices index suggests an improvement of 0.28 standard deviation for treat- ment mentees, with the largest gain observed for marketing practices (a 0.27 SD improvement over control mentors). Record keeping, stock control and financial planning, similarly, show improvements in the range of 0.18 and 0.27 SD for treatment mentees compared to pure control mentors. The magnitude of the impact on ‘spillover’ mentees is similar in size to the treatment mentees; in fact, we fail to reject the hypothesis that the coefficients for treatment mentees and spillover mentees are equal for indices of the four different types of business practices (although, as one might expect, the point estimates are lower). These results indicate widespread spillover impacts among the control mentees of treated mentors. In Table 8, we see a (marginally) statistically significant impact of being mentored on the index of business expansion, which is comprised of the total number of businesses that the mentees have, as the number of unique products that they sell and their number of employees. There is no significant impact on the spillover mentees, which would be consistent either with the lower point estimates of changes in skills or that the mentoring had a wider range of impacts when it was a direct, rather than indirect, relationship. Taken together these results show that the treatment impacted the practices, but apparently not the profits of the mentees of the treated mentors. This is counter to the results of the mentors, who show a change in both practices and profits. The lack of effect of being mentored on the business profits could possibly be attributed to the fact that it was unable to induce large enough changes in the business practices of the mentees. This is consistent with Mckenzie and Woodruff’s (2016) argument that the relationship between ‘better’ business practices and performance is causal and that the lack of effect for existing business trainings is because they do not change business practices enough rather than because these practices do not matter. Along the lines of Mckenzie and Woodruff (2016), we investigate whether the effect of the mentoring treatment on profit is consistent with the predicted effect given the observed changes in business practices, to which we now turn. Business practices and profits for mentees - Power issues The point estimate for the impact of mentoring on winsorized profits as seen in Col (3) of Table 6, is 216 Ethiopian Birr (ETB), which is 14.4% of the (pure) control mean. A 95% 18 confidence interval is [-97 ETB, + 533 ETB] or [-6%, + 35%] which allows for mentoring to have had a relatively sizable impact on profits; just that there is no power to detect them. Among the (pure) control mentees, a positive and statistically significant correlation between business practices and business outcomes is observed. Table 9 shows that a 1 SD increase in the business practice score is associated with an improvement in (winsorized) profit by 822 ETB7 . If we multiply this by the mentoring treatment effect on business practices in Table 7 (mentoring causes an improvement of business practice index by 0.27 SD) we get an estimate of how much the change in business practices is likely to change profits, which is 0.27 x 828 ETB = 223.6 ETB. Not only is this within the confidence interval of the estimate of profits, it’s also quite close to the point estimate of 216 ETB for the mentoring treatment effect on business profit. 6 Conclusion The Women in Agribusiness Leaders Network (WALN) project is an interesting departure from the standard business training model. Under a stratified randomized controlled trial framework, experienced business women who were involved in agriculture-related businesses were given formal business training. The more innovative part in this project is a second- stage mentoring component in which a random selection of women mentees within the social and business network of mentors received customized mentoring from the trained mentors. Over a post-training period of three years, we find that formal business training for the mentors causes large impacts on business performance such as reported profits as well as the number of business activities. This effect appears to be driven by the continued adoption of several of the ‘better’ business practices identified by McKenzie and Woodruff (2016). Given the extent of micro-entrepreneurial activities in developing countries, it is also im- portant to identify low-cost interventions that can lead to the adoption of these business practices that may improve performance for small businesses. Mentoring could be a rela- tively low-cost solution to this problem. We tested this by connecting trainees of the formal business training to a random subset of small firm owners within their social and business networks and evaluate the outcome of the treatment mentees compared to control mentees 7 McKenzie and Woodruff (2016) report an increase of profits in the range of 65% to 100% for a 1 SD improvement in business practices score. Among the pure control mentees in our sample, this association is at 91% (Table 9). This comes with a caveat - we use IHS transformations whereas McKenzie and Woodruff (2016) use log transformations for profits. 19 (who did not receive the mentoring). We find that the treatment mentees exhibit strong effects in the adoption of some business practices; however, this does not translate to higher profits. The apparent muted results for the impact of mentoring on profits are consistent with the lackluster results from recent evaluations of business mentoring such as Brooks, Donovan, and Johnson (2018). It is, however, possible that the adoption of practices observed among treatment mentees is not of a large enough magnitude to change profits substantially. We find some evidence for this possibility—the effect of mentoring on profit is consistent with the predicted effect given the observed changes in business practices. This is similar to the exercise carried out by McKenzie and Woodruff (2016), in which they show that the modest observed effects of business training on profits and sales in six developing countries are remarkably consistent with the predicted effects given the observed changes in business practices. 7 Tables 20 Table 1: Mentor Baseline Balance (comparing Treatment vs. Control Mentors) (1) (2) T-test Control Mentor Treatment Mentor Difference Variable N Mean/SE N Mean/SE (1)-(2) Monthly Profit (Raw) 76 17982.593 83 1.04e+06 -1.02e+06 (10186.831) (1.00e+06) Monthly Profit (Winsorized) 76 5751.117 83 5816.470 -65.353 (1194.538) (1036.791) Monthly Revenue (Raw) 96 72306.990 97 42861.835 29445.155 (52534.630) (17733.741) Monthly Revenue (Winsorized) 96 14171.573 99 15834.323 -1662.750 (2649.646) (2982.843) Number of products 98 5.531 99 5.556 -0.025 (0.840) (0.892) Total Num. of Businesses 98 1.531 99 1.687 -0.156 (0.076) (0.090) Owner’s age 98 37.500 99 37.424 0.076 (0.919) (0.924) Owner’s Years of School 98 9.214 99 9.808 -0.594 (0.525) (0.488) Operational years of business 96 6.823 94 7.787 -0.964 (0.519) (0.634) Owner’s years of bus. experience 98 11.031 99 11.697 -0.666 (0.797) (0.800) Baseline Prac. Score (%) 98 30.952 99 39.057 -8.105 (3.495) (3.889) Has Bus. Plan 98 31.633 99 40.404 -8.771 (4.722) (4.957) Has Annual Budget 98 22.449 99 30.303 -7.854 (4.236) (4.642) Has Financial Record 98 38.776 99 46.465 -7.689 (4.947) (5.038) Notes : The value displayed for t-tests are the differences in the means across the groups. Fixed effects using stratas are included in all estimation regressions. ***, **, and * indicate significance at the 1, 5, and 10 percent critical level. 21 Table 2: Mentee Baseline Balance (comparing Pure Control, Spillover and Treatment Mentees) (1) (2) (3) T-test Pure Contr. mentee Spillover Contr. Mentee Treatment Mentee Difference Variable N/[Clusters] Mean/SE N/[Clusters] Mean/SE N/[Clusters] Mean/SE (1)-(2) (1)-(3) (2)-(3) Monthly Profit (Raw) 531 2433.547 283 61300.992 283 10409.782 -5.89e+04 -7976.235 50891.210 [90] (510.955) [94] (58946.793) [94] (7164.466) Monthly Profit (Winsorized) 531 1374.332 283 1657.882 283 1712.992 -283.550 -338.660* -55.110 [90] (124.080) [94] (142.270) [94] (164.861) Monthly Revenue (Raw) 531 15236.655 291 17223.773 292 42132.318 -1987.118 -2.69e+04 -2.49e+04 [90] (6666.054) [94] (5105.080) [94] (23805.750) Monthly Revenue (Winsorized) 539 5484.071 294 6854.976 295 6703.041 -1370.906 -1218.970 151.936 [90] (497.579) [94] (695.983) [94] (644.533) Number of products 539 4.824 293 5.604 295 5.003 -0.780 -0.180 0.601 [90] (0.510) [94] (0.708) [94] (0.550) Total Num. of Businesses 539 1.468 294 1.374 295 1.393 0.093 0.074 -0.019 [90] (0.038) [94] (0.043) [94] (0.044) Owner’s age 538 35.626 293 35.188 295 34.786 0.439 0.840 0.401 [90] (0.586) [94] (0.784) [94] (0.685) Owner’s Years of School 539 7.492 294 7.694 295 8.244 -0.202 -0.752* -0.550 [90] (0.361) [94] (0.376) [94] (0.319) Operational years of business 493 7.473 267 6.719 271 7.018 0.754 0.454 -0.299 [90] (0.713) [92] (0.551) [93] (0.497) Owner’s years of bus. experience 539 10.477 294 9.918 295 9.793 0.558 0.684 0.125 [90] (0.706) [94] (0.646) [94] (0.554) Baseline Prac. Score (%) 539 12.554 294 14.286 295 18.644 -1.732 -6.090* -4.358* [90] (1.647) [94] (1.992) [94] (2.457) Has Bus. Plan 539 11.317 294 14.286 295 17.627 -2.968 -6.310 -3.341 [90] (1.962) [94] (2.430) [94] (2.821) Has Annual Budget 539 8.905 294 6.463 295 12.881 2.443 -3.976 -6.419*** [90] (1.595) [94] (1.910) [94] (2.583) Has Financial Record 539 17.440 294 22.109 295 25.424 -4.669 -7.984** -3.315 [90] (2.220) [94] (2.885) [94] (2.997) Notes : The value displayed for t-t ests are the differences in the means across the groups. Fixed effects using stratas are included in all estimation regressions. Standard errors are clustered at the mentor level. ***, **, and * indicate significance at the 1, 5, and 10 percent critical level. 22 Table 3: Reduced Form Effects of Business Training on Last Month’s Profits and Revenue (1) (2) (3) (4) (5) (6) (7) Profit Profit All Profit Revenue Revenue Revenue Profit & (Raw) (IHS) (Wins.) (Raw) (IHS) (Wins.) Rev. Index Bus. Training -6845.05 1.03∗∗ 2402.34∗∗∗ 16493.46 0.71∗ 10282.90∗∗ 0.21∗∗∗ (17788.73) (0.44) (689.06) (14575.03) (0.36) (4020.05) (0.06) Omitted category mean 22940.79 5.90 2985.30 31221.12 8.26 16530.93 -0.12 Number of firms 519 519 519 507 507 507 519 Number of distinct firms 193 193 193 191 191 191 193 Notes: The outcome variable in Col (7) is the mean of the standardized z-scores of various profits and revenue measures (raw, IHS and winsorized) for the last 30 days. 3 rounds of follow-up data pooled. Strata fixed effects are used in each estimation. SEs are clustered at firm level. 23 Table 4: Reduced Form Effects of Business Training on Bus. Practices (1) (2) (3) (4) (5) Marketing Stock Control Recordkeeping Financial Planning Bus. Prac. Index Index Index Index Index Mean: (1)-(4) Bus. Training 0.139∗∗∗ 0.082∗∗ 0.125∗∗∗ 0.076∗∗ 12.304∗∗∗ (0.040) (0.040) (0.042) (0.035) (3.393) Omitted category mean 0.381 0.584 0.516 0.393 39.047 Number of firms 394 327 327 327 394 Number of distinct firms 197 177 177 177 197 Notes: The last 2 rounds of follow-up data have been pooled for these estimations. Strata fixed effects are used in each estimation. SEs are clustered at firm level. The outcome variable in Col. (1) is the proportion of marketing practices used by business; Col. (2) is the proportion of buying and stock control practices; Col. (3) is the proportion of recordkeeping practices; Col. (4) is the proportion of financial planning practices used. The outcome variable in Col. (5) is the proportion of 26 business practices (in marketing, stockkeeping, recordkeeping, and financial planning) identified by Mckenzie and Woodruff (2016) which predict higher survival rates and growth of businesses. 24 Table 5: Reduced Form Effects of Business Training on Business Expansion (1) (2) (3) (4) Total No. of No. of No. of Index of Bus. Expansion Businesses Products Employees Mean: (1)-(4) Bus. Training 0.226∗∗ 0.999 0.383 0.129 (0.101) (1.150) (1.664) (0.080) Omitted category mean 1.626 5.458 7.601 -0.118 Number of firms 530 349 349 360 Number of distinct firms 194 191 191 194 Notes: 3 rounds of follow-up data pooled. Strata fixed effects are used in each estimation. SEs are clustered at firm level. 25 Table 6: Reduced Form Effects of Receiving Mentorship on Last Month’s Profits and Revenue (1) (2) (3) (4) (5) (6) (7) Profit Profit All Profit Revenue Revenue Revenue Profit & (Raw) (IHS) (Wins.) (Raw) (IHS) (Wins.) Rev. Index Spillover Contr. Mentee -3176.404 0.067 107.119 -42218.547 0.308 884.848 0.026 (4381.110) (0.248) (155.578) (45230.866) (0.245) (876.550) (0.046) Treatment Mentee -4650.870 0.234 215.606 -41490.286 0.533∗∗ 1186.038 0.065 (5952.535) (0.240) (158.598) (58726.659) (0.247) (877.064) (0.052) Treatment – Spillover -1.5e+03 0.167 108.487 728.261 0.225 301.191 0.039 p-value 0.528 0.464 0.478 0.969 0.238 0.698 0.375 Omitted category mean 4869.178 5.989 1512.760 4.2e+04 7.616 7065.762 -0.039 Number of firms 2,774 2,774 2,774 2,678 2,678 2,678 2,774 Number of distinct firms 1,051 1,051 1,051 1,028 1,028 1,028 1,051 26 Notes: The outcome variable in Col (7) is the mean of the standardized scores of various profits and revenue measures (raw, IHS and winsorized profits and revenue) for the last 30 days. The ommitted category in all the specifications is pure control mentee. 2 rounds of follow-up data are pooled. Strata fixed effects are used in each estimation. SEs are clustered at mentor level. Table 7: Reduced Form Effects of Receiving Mentorship on Bus. Practices (Standardized Z-scores) (1) (2) (3) (4) (5) Marketing Stock Control Recordkeeping Financial Planning Bus. Prac. Index Index Index Index Index Mean: (1)-(4) Spillover Contr. Mentee 0.209∗∗∗ 0.162∗∗ 0.144∗∗ 0.173∗∗ 0.182∗∗ (0.073) (0.074) (0.072) (0.071) (0.074) Treatment Mentee 0.265∗∗∗ 0.182∗∗∗ 0.179∗∗ 0.265∗∗∗ 0.270∗∗∗ (0.075) (0.069) (0.077) (0.080) (0.075) Treatment – Spillover 0.056 0.020 0.035 0.092 0.088 p-value 0.429 0.742 0.610 0.237 0.205 Omitted category mean -0.126 -0.090 -0.098 -0.124 -0.124 Number of firms 2,256 1,800 1,800 1,800 2,256 Number of distinct firms 1,126 979 979 979 1,126 27 Notes: The outcome variable in each column has been standardized. The outcome variable in Col (1) is the proportion of marketing practices used by business; Col (2) is the proportion of buying and stock control practices; Col (3) is the proportion of recordkeeping practices; Col (4) is the proportion of financial planning practices used. The outcome variable in Col (5) is the proportion of 26 business practices (in marketing, stockkeeping, recordkeeping, and financial planning) identified by Mckenzie and Woodruff (2016) which predict higher survival rates and growth of businesses. The ommitted category in all the specifications is pure control mentee. 2 rounds of follow-up data are pooled. Strata fixed effects are used in each estimation. SEs are clustered at mentor level. Table 8: Reduced Form Effects of Receiving Mentorship on No. of Businesses/IGAs (1) (2) (3) (4) Total No. of No. of No. of Index of Bus. Expansion Businesses Products Employees Mean: (1)-(4) Spillover Contr. Mentee -0.024 -0.188 -0.081 0.040 (0.060) (0.524) (0.470) (0.048) Treatment Mentee 0.088 -0.285 0.372 0.094∗ (0.062) (0.445) (0.874) (0.049) Treatment – Spillover 0.111 -0.097 0.453 0.054 p-value 0.042 0.821 0.548 0.296 Omitted category mean 1.452 5.032 3.580 -0.114 Number of firms 2,931 1,878 1,772 1,984 Number of distinct firms 1,084 1,044 979 1,084 28 Notes: ANCOVA specification for Col. 2. 3 rounds of follow-up data pooled. Strata fixed effects are used in each estimation. SEs are clustered at mentor level. Table 9: Correlation of Bus. Practices and Outcomes (Revenue/Profit) for (pure) Control Mentees (1) (2) (3) Profit & All Profit All Profit Rev. Index (Wins.) (IHS) Bus. Prac. Score (Z-Score) 0.27∗∗∗ 821.54∗∗∗ 0.91∗∗∗ (0.03) (118.70) (0.23) Number of firms 845 845 845 29 Number of distinct firms 463 463 463 Notes: 2 rounds of follow-up data pooled. The outcome variable in Col (1) is the mean of the standardized z-scores of various profits and revenue measures (raw, IHS and winsorized profits and revenue) for the last 30 days. Column (2) is winsorized monthly profit. For Column (3), the outcome variable is IHS-transformed monthly profit. References Allen, Tammy D., Lillian T. Eby, Mark L. Poteet, Elizabeth Lentz, and Lizzette Lima (2004). “Career Benefits Associated With Mentoring for Proteges: A Meta-Analysis”. Journal of Applied Psychology 89.1. Place: US Publisher: American Psychological Association, pp. 127–136. doi: 10.1037/0021- 9010.89.1.127. Atkin, David, Amit K. Khandelwal, and Adam Osman (May 2017). “Exporting and Firm Perfor- mance: Evidence from a Randomized Experiment”. The Quarterly Journal of Economics 132.2. Publisher: Oxford Academic, pp. 551–615. doi: 10.1093/qje/qjx002. Bandiera, Oriana and Imran Rasul (Oct. 2006). “Social Networks and Technology Adoption in North- ern Mozambique”. The Economic Journal 116.514. Publisher: Oxford Academic, pp. 869–902. doi: 10.1111/j.1468-0297.2006.01115.x. Beaman, Lori, Ariel BenYishay, Jeremy Magruder, and Ahmed Mushfiq Mobarak (Aug. 2018). “Can Network Theory-based Targeting Increase Technology Adoption?” Working Paper 24912. Series: Working Paper Series. National Bureau of Economic Research. doi: 10.3386/w24912. BenYishay, Ariel and A. Mushfiq Mobarak (May 2019). “Social Learning and Incentives for Ex- perimentation and Communication”. The Review of Economic Studies 86.3. Publisher: Oxford Academic, pp. 976–1009. doi: 10.1093/restud/rdy039. Blattman, Christopher and Laura Ralston (July 2015). “Generating Employment in Poor and Fragile States: Evidence from Labor Market and Entrepreneurship Programs”. SSRN Scholarly Paper ID 2622220. Rochester, NY: Social Science Research Network. doi: 10.2139/ssrn.2622220. Bloom, Nicholas, Benn Eifert, Aprajit Mahajan, David McKenzie, and John Roberts (Feb. 2013). “Does Management Matter? Evidence from India”. The Quarterly Journal of Economics 128.1. Publisher: Oxford Academic, pp. 1–51. doi: 10.1093/qje/qjs044. Bloom, Nicholas and John Van Reenen (Nov. 2007). “Measuring and Explaining Management Prac- tices Across Firms and Countries”. The Quarterly Journal of Economics 122.4. Publisher: Oxford Academic, pp. 1351–1408. doi: 10.1162/qjec.2007.122.4.1351. Brooks, Wyatt, Kevin Donovan, and Terence R. Johnson (Oct. 2018). “Mentors or Teachers? Mi- croenterprise Training in Kenya”. American Economic Journal: Applied Economics 10.4, pp. 196– 221. doi: 10.1257/app.20170042. Bruhn, Miriam, Dean Karlan, and Antoinette Schoar (May 2010). “What Capital is Missing in Developing Countries?” American Economic Review 100.2, pp. 629–633. doi: 10.1257/aer.100. 2.629. – (Apr. 2018). “The Impact of Consulting Services on Small and Medium Enterprises: Evidence from a Randomized Trial in Mexico”. Journal of Political Economy 126.2. Publisher: The University of Chicago Press, pp. 635–687. doi: 10.1086/696154. Burke, R. J. and C. A. McKeen (Apr. 1990). “Mentoring in organizations: Implications for women”. Journal of Business Ethics 9.4, pp. 317–332. doi: 10.1007/BF00380330. Cai, Jing and Adam Szeidl (Aug. 2018). “Interfirm Relationships and Business Performance”. The Quarterly Journal of Economics 133.3, pp. 1229–1282. doi: 10.1093/qje/qjx049. 30 Calderon, Gabriela, Jesse M. Cunha, and Giacomo De Giorgi (Jan. 2020). “Business Literacy and Development: Evidence from a Randomized Controlled Trial in Rural Mexico”. Economic Devel- opment and Cultural Change 68.2. Publisher: The University of Chicago Press, pp. 507–540. doi: 10.1086/701213. Campos, Francisco, Michael Frese, Markus Goldstein, Leonardo Iacovone, Hillary C. Johnson, David McKenzie, and Mona Mensmann (Sept. 2017). “Teaching personal initiative beats traditional train- ing in boosting small business in West Africa”. Science 357.6357. Publisher: American Association for the Advancement of Science Section: Report, pp. 1287–1290. doi: 10.1126/science.aan5329. Campos, Francisco, Markus Goldstein, Laura McGorman, Ana Maria Munoz Boudet, and Obert Pimhidzai (Dec. 2015). Breaking the Metal Ceiling: Female Entrepreneurs Who Succeed in Male- Dominated Sectors. Policy Research Working Papers. The World Bank. doi: 10.1596/1813-9450- 7503. Conley, Timothy G. and Christopher R. Udry (Mar. 2010). “Learning about a New Technology: Pineapple in Ghana”. American Economic Review 100.1, pp. 35–69. doi: 10.1257/aer.100.1.35. David McKenzie Susana Puerto (Mar. 2017). Growing Markets through Business Training for Fe- male Entrepreneurs: A Market-Level Randomized Experiment in Kenya. Policy Research Working Papers. The World Bank. doi: 10.1596/1813-9450-7993. De Mel, Suresh, David McKenzie, and Christopher Woodruff (July 2012). Business Training and Female Enterprise Start-up, Growth, and Dynamics: Experimental Evidence from Sri Lanka. Policy Research Working Papers. The World Bank. doi: 10.1596/1813-9450-6145. Drexler, Alejandro, Greg Fischer, and Antoinette Schoar (Apr. 2014). “Keeping It Simple: Financial Literacy and Rules of Thumb”. American Economic Journal: Applied Economics 6.2, pp. 1–31. doi: 10.1257/app.6.2.1. Foster, Andrew D. and Mark R. Rosenzweig (Dec. 1995). “Learning by Doing and Learning from Others: Human Capital and Technical Change in Agriculture”. Journal of Political Economy 103.6. Publisher: The University of Chicago Press, pp. 1176–1209. doi: 10.1086/601447. Gine, Xavier and Ghazala Mansuri (June 2014). Money or Ideas? A Field Experiment on Constraints to Entrepreneurship in Rural Pakistan. Policy Research Working Papers. The World Bank. doi: 10.1596/1813-9450-6959. Ginther, Donna K, Janet Currie, Francine D Blau, and Rachel Croson (Mar. 2020). “Can Mentoring Help Female Assistant Professors in Economics? An Evaluation by Randomized Trial”. Working Paper 26864. Series: Working Paper Series. National Bureau of Economic Research. doi: 10.3386/ w26864. Hansford, Brian, Lee Tennent, and Lisa Catherine Ehrich (Aug. 2002). “Business Mentoring: Help or hindrance?” Mentoring & Tutoring: Partnership in Learning 10.2. Publisher: Routledge _eprint: https://doi.org/10.1080/1361126022000002428, pp. 101–115. doi: 10.1080/1361126022000002428. Higuchi, Yuki, Edwin P. Mhede, and Tetsushi Sonobe (Feb. 2019). “Short- and medium-run impacts of management training: An experiment in Tanzania”. World Development 114, pp. 220–236. doi: 10.1016/j.worlddev.2018.10.002. Jackson, Matthew O. (Nov. 2010). Social and Economic Networks. Google-Books-ID: rFzHinVAq7gC. Princeton University Press. 31 St-Jean, Etienne (2012). “Mentoring as professional development for novice entrepreneurs: maximiz- ing the learning1”. International Journal of Training and Development 16.3. _eprint: https://onlinelibrary.wil 2419.2012.00404.x, pp. 200–216. doi: 10.1111/j.1468-2419.2012.00404.x. St-Jean, Etienne and Josee Audet (Mar. 2012). “The role of mentoring in the learning development of the novice entrepreneur”. International Entrepreneurship and Management Journal 8.1, pp. 119– 140. doi: 10.1007/s11365-009-0130-7. Kandpal, Eeshani, Kathy Baylis, and Mary Arends-Kuenning (Feb. 2012). “Empowering Women Through Education and Influence: An Evaluation of the Indian Mahila Samakhya Program”. SSRN Scholarly Paper ID 2010942. Rochester, NY: Social Science Research Network. https://papers. ssrn.com/abstract=2010942. Karlan, Dean, Ryan Knight, and Christopher Udry (Oct. 2015). “Consulting and capital experiments with microenterprise tailors in Ghana”. Journal of Economic Behavior & Organization. Economic Experiments in Developing Countries 118, pp. 281–302. doi: 10.1016/j.jebo.2015.04.005. Karlan, Dean and Martin Valdivia (Aug. 2010). “Teaching Entrepreneurship: Impact of Business Training on Microfinance Clients and Institutions”. The Review of Economics and Statistics 93.2. Publisher: MIT Press, pp. 510–527. doi: 10.1162/REST_a_00074. Kram, Kathy E. (Dec. 1983). “Phases of the Mentor Relationship”. Academy of Management Journal 26.4. Publisher: Academy of Management, pp. 608–625. doi: 10.5465/255910. – (1988). Mentoring at work: Developmental relationships in organizational life. Mentoring at work: Developmental relationships in organizational life. Pages: xiii, 252. Lanham, MD, England: Uni- versity Press of America. Lafortune, Jeanne, Julio Riutort, and Jose Tessada (Oct. 2018). “Role Models or Individual Consult- ing: The Impact of Personalizing Micro-entrepreneurship Training”. American Economic Journal: Applied Economics 10.4, pp. 222–245. doi: 10.1257/app.20170077. McKenzie, David (Nov. 2012). “Beyond baseline and follow-up: The case for more T in experiments”. Journal of Development Economics 99.2, pp. 210–221. doi: 10.1016/j.jdeveco.2012.01.002. McKenzie, David and Christopher Woodruff (Feb. 2014). “What Are We Learning from Business Training and Entrepreneurship Evaluations around the Developing World?” The World Bank Re- search Observer 29.1. Publisher: Oxford Academic, pp. 48–82. doi: 10.1093/wbro/lkt007. – (July 2016). “Business Practices in Small Firms in Developing Countries”. Management Science 63.9, pp. 2967–2981. doi: 10.1287/mnsc.2016.2492. Munshi, Kaivan (Feb. 2004). “Social learning in a heterogeneous population: technology diffusion in the Indian Green Revolution”. Journal of Development Economics 73.1, pp. 185–213. doi: 10.1016/j.jdeveco.2003.03.003. Nelson, Debra L. and James C. Quick (Apr. 1985). “Professional Women: Are Distress and Disease Inevitable?” Academy of Management Review 10.2. Publisher: Academy of Management, pp. 206– 218. doi: 10.5465/amr.1985.4277941. Noe, Raymond A. (Jan. 1988). “Women and Mentoring: A Review and Research Agenda”. Academy of Management Review 13.1. Publisher: Academy of Management, pp. 65–78. doi: 10.5465/amr. 1988.4306784. 32 Quinn, Simon and Christopher Woodruff (2019). “Experiments and Entrepreneurship in Developing Countries”. Annual Review of Economics 11.1, pp. 225–248. doi: 10.1146/annurev-economics- 080218-030246. Valdivia, Martin (Mar. 2015). “Business training plus for female entrepreneurship? Short and medium- term experimental evidence from Peru”. Journal of Development Economics 113, pp. 33–51. doi: 10.1016/j.jdeveco.2014.10.005. 33 A Appendix Table A.1: Number of Firms by Survey Rounds and Treatment Status Treatment Type Baseline Follow-up 1 Follow-up 2 Follow-up 3 Control Mentor 98 96 91 89 Treatment Mentor 99 97 95 95 Total - Mentors 197 193 186 184 (Pure) Control Mentee 539 512 501 500 Spillover Mentee 294 277 270 276 Treatment Mentee 295 285 278 282 Total - Mentees 1,128 1,074 1,049 1,058 34 Table A.2: WALN Selection Criteria Common criteria for mentors and mentees: • Must have business operations within the geographical scope of the project, i.e. the 96 AGP woredas. • Must be active in one of the targeted value chains (chickpeas, coffee, honey, maize, sesame, wheat, livestock and dairy). This can be any part of the value chain, such as production, processing, trading, transport, agricultural finance or other services. Specific criteria for mentors and mentees: Mentors Mentees • Must be an owner and/or manager of a • Must be recommended by a mentor who registered business or association believes that the mentee has the ability to learn and contribute positively to their sector and to WALN • Must have an individual or business bank • Must own, manage and/or hold a senior account and email address* position in a registered business or association* • Must be interested in taking part in • Must be located geographically close training sessions to develop their skills in enough to the mentor to participate in financial planning, management, marketing, monthly meetings negotiation, public speaking as well as meeting deliverables that demonstrate a willingness and dedication to put in the effort to change behaviors and learn new ways to operate • Must be willing to mentor at least 3 • Must be willing and able to attend aspiring businesswomen in targeted regions mentoring sessions over a six-month period through monthly group sessions as well as one-on-one follow up • Must be a recognized leader in the • Must be open to change behaviors and industry and/or community learn new ways to operate business • Must sign a commitment form • Must be open to become a new mentor for aspiring businesswomen in targeted regions in the future *Some flexibility in judging compliance with those criteria may be applied. In particular, women whose business is based in Addis-Ababa or in any major city of the targeted regions (capital city or trade town such as Gondar, Debre Markos or Adama) may be considered for the program, given the value-addition, in terms of market linkage, they would bring to the network. 35 Table A.3: Modules Covered in Business Training Module Title Module Description Business Development and Planning Introduction to business planning, including an overview of how to evaluate business success, business planning, tools for marketing management, financial management and analysis, stock and value chain management and human resource management. (~31 hours) Leadership Building women’s confidence in leadership roles through practical managerial skills, team building, decision making, conflict management and leadership styles. There was a special focus on the role of gender in leadership. Participants also received leadership training from qualified specialists. (~12 hours) Communication and Networking This module was focussed on improving communication within the mentor’s own business. Improving communication in the business context by highlighting the importance of communication, the role it plays in business success, particularly focussing on skills to overcome barriers to effective communication. (~9 hours) Mentoring and Coaching Specific knowledge and practical skills on the value and process of mentoring other women entrepreneurs enhance their professional and personal competencies to succeed as businesswomen. (~6 hours) 36 Table A.4: Reduced Form Effects of Business Training on Survey Response Rates - Trajectory (1) (2) (3) Follow-up 1 Follow-up 2 Follow-up 3 Bus. Training -0.00 0.03 0.05 (0.02) (0.03) (0.04) Omitted category mean 0.98 0.93 0.91 Number of firms 197 197 197 Number of distinct firms 197 197 197 Number of strata 18 18 18 Notes: Regressions for 3 rounds of follow-up data separately carried out. Strata Fixed effects. SEs not clustered. Table A.5: Reduced Form Effects of Receiving Mentorship on Survey Response Rates - Trajectory (1) (2) (3) Follow-up 1 Follow-up 2 Follow-up 3 Spillover Contr. mentee -0.007 -0.011 0.012 (0.016) (0.021) (0.024) Treatment mentee 0.016 0.010 0.029 (0.015) (0.018) (0.022) Treatment – Spillover 0.022 0.022 0.017 p-value 0.137 0.273 0.425 Omitted category mean 0.950 0.929 0.928 Number of firms 1,128 1,128 1,128 Number of distinct firms 1,126 1,126 1,126 Notes: Regressions for 3 rounds of follow-up data separately carried out. The ommitted category in all the specifications is pure control mentee. Strata fixed effects are used in each estimation. SEs are clustered at the mentor level. 37 Figure A.1: Cumulative Density Function (CDF) of Profit and Revenue Indices - by Treatment Status of Mentors 1 Cumul. Density 0 -2 0 2 4 Avg. Z-score for Last Month's Profit and Revenue (Follow-up 3 data) 0. Control Mentor 1. Treatment Mentor 1 Cumul. Density 0 -2 0 2 4 6 8 Avg. Z-score for Last Month's Profit and Revenue (Follow-up 2 data) 0. Control Mentor 1. Treatment Mentor A.1 Outcomes of interest • Profits (and revenue) index: This index averaged the z-scores of the following variables accord- ing to the procedure described above: – Raw last month’s profits: We asked business owners for profits directly rather than at- tempting to have them match up revenue and expenses for the same period. Entrepreneurs were asked: “What were the profits of all your businesses in the last full month? That is, your revenue after having paid all expenses including the salaries of employees, but before paying your own salary?”8 – Last month’s profits, transformed using the inverse hyperbolic sine – Last month’s profits, winsorized9 at the 95th and 5th percentiles – Raw last month’s revenue: Entrepreneurs were asked “what was the revenue of your business in the last full month?” – Last month’s revenue, transformed using the inverse hyperbolic sine – Last month’s revenue, winsorized at the 95th percentile The outcome variables in Tables 4 and 7 are indices of 26 business practices for marketing prac- tices, record-keeping, stock control and financial planning. For most of the business practices, the entrepreneur was asked whether they used the practice over the previous 3 months. These practices were recoded to binary, with those saying they did not use the practice in the previous 3 months being coded to 0. The business practices are defined in the following way. • Index for Marketing and customer service practices: This index averaged the z-scores of the following variables – Visited competitor’s business to see prices – Visited competitor’s business to see products – Asked existing customers what other products they should offer – Talked with former customer to see why stopped buying – Asked supplier what products selling well – Used a special offer to attract customers – Have done advertising in last 6 months- constructed from a question asking about whether they used the following forms of advertising: ∗ Written press ∗ Radio or television ∗ Classified ads through professional, trade or religious associations ∗ Trade fair ∗ Posters/flyers/business cards 8 Note that the profits of all businesses for the previous month is not available for the baseline survey, rather the profit for all businesses in the previous one year was asked. We divide this number by 12 to derive monthly estimates. 9 Winsorization was done separately for the mentors and mentees’ distributions 39 • Record keeping practices index – Keep written records – Record every purchase and sale – Can use records to know cash on hand – Use records to know whether sales of product increase or decrease – Worked out cost of each main product – Know which goods make most profit per item – Have a written budget for monthly expenses – Have records that could document ability to pay to bank • Stock Control index – Negotiated price with supplier in the last 3 months – Compared price with other supplier in the last 3 months – Don’t run out of stock frequently ∗ this is measured from the question on how often the business runs out of stock; any time interval less than 6 months is coded as 1 and the rest as 0 • Financial Planning index – review financial performance monthly – have sales target for next year – compare sales goal to target monthly – have a budget of costs for next year – prepare profit and loss statement – prepare cashflow statement – prepare balance sheet – prepare income and expenditure statement Data on the following business practices are collected in every round:10 – Has a written business plan – Has a written annual budget – Keeps financial record for each transaction The outcome variables underlying the business diversification and expansion index in Tables 5 and 8 are defined in the following way: 10 Data on the 26 business practices for marketing practices, record-keeping, stock control and financial planning have been collected in the last 2 follow-up rounds of data collection. On the other hand, these 3 questions on written business plan, written annual budget and whether the firm keeps financial record for each transaction have been asked in every round of data collection. 40 • Number of businesses: total number of independent income-generating activities the business- owner has; this offers a measure of business diversification • Total number of products: this denotes the total number independent products the business sells • Total number of employees 41 A.2 Business Practices Table A.6: Reduced Form Effects of Business Training on Marketing Practices (1) (2) (3) (4) (5) (6) (7) Visited Visited Asked Asked Frmr Asked Supplier Attracted Any Competitors Competitors Customers Customer Product Customer Advertise Compare Prices Compare Prod. Prod Needs why Left Sell Well Sp Offers 6 mnths Bus. Training 0.113∗∗ 0.133∗∗ 0.057 0.204∗∗∗ 0.194∗∗∗ 0.172∗∗∗ 0.020 (0.053) (0.058) (0.053) (0.077) (0.050) (0.062) (0.044) Omitted category mean 0.691 0.642 0.690 0.544 0.240 0.394 0.174 Number of firms 318 315 327 202 394 327 327 Number of distinct firms 177 177 177 141 197 177 177 Notes: The last 2 rounds of follow-up data have been pooled for these estimations. Strata fixed effects are used in each estimation. SEs are clustered at firm level. The outcome variable in Col. (1) indicates whether the business visited at least one of its competitor’s businesses to see what prices its competitors are charging; Col. (2): Visited at least one of its competitor’s businesses to see what products its competitors have available for sale; Col. (3): Asked existing customers whether there are any other products the customers would like the business to sell or produce; Col. (4): Talked with at least one former customer to find out why former customers have stopped buying from this business; Col. (5): Asked a supplier about which products are selling well in this business industry; Col. (6): Attracted customers with a special offer ; Col. (7): Advertised in any form (last 6 months). Table A.7: Reduced Form Effects of Business Training on Buying and Stock Control Practices (1) (2) (3) Negotiated Compared Doesn’t or Attempted Price/Quality Run Out w. Suppliers Alt. Suppliers of Stock Bus. Training 0.161∗∗∗ 0.082 0.022 (0.056) (0.054) (0.069) Omitted category mean 0.587 0.665 0.437 Number of firms 327 327 231 Number of distinct firms 177 177 164 Notes: The last 2 rounds of follow-up data have been pooled for these estimations. Strata fixed effects are used in each estimation. SEs are clustered at firm level. The outcome variables in Col. (1) indicates whether the business attempted to negotiate with a supplier for a lower price on raw material; Col. (2): Compared the prices or quality offered by alternate suppliers or sources of raw materials to the business’ current suppliers or sources of raw 43 material; Col. (3): Does not run out of stock monthly or more. Table A.8: Reduced Form Effects of Business Training on Record-keeping Practices (1) (2) (3) (4) (5) (6) (7) (8) Keeps Records Use Records Use Records Works out Knows Has Has Records Written Bus. Every to Calcul. Evaluate Cost to Bus Most Profitable Written Bank Loan Records Purchase/Sale Cash on Hand Prod Sale of each Prod Goods Budget Purposes Bus. Training 0.214∗∗∗ 0.195∗∗∗ 0.185∗∗∗ 0.209∗∗∗ 0.035 0.049 0.014 0.080 (0.060) (0.061) (0.060) (0.059) (0.040) (0.032) (0.053) (0.071) Omitted category mean 0.432 0.426 0.419 0.400 0.839 0.890 0.310 0.430 Number of firms 327 327 327 327 327 327 327 251 Number of distinct firms 177 177 177 177 177 177 177 167 Notes: The last 2 rounds of follow-up data have been pooled for these estimations. Strata fixed effects are used in each estimation. SEs are clustered at firm level. The outcome variable in Col. (1) indicates whether the business keeps written business records; Col. (2): Records every purchase and sale made by the business; Col. (3): Able to use records to see how much cash the business has on hand at any point in time; Col. (4): Uses records regularly to know whether sales of a particular product are increasing or decreasing from one month to another; Col. (5): Works out the cost to the business of each main product it sells; Col. (6): Knows which goods you make the most profit per item selling; Col. (7): Has a written budget, which states how much is owed each month for rent, electricity, equipment maintenance, transport, advertising, and other indirect costs to business; Col. (8): Has records documenting that there exists enough money each month after paying businessexpenses to repay a loan in the hypothetical situation that this business wants a bank loan Table A.9: Reduced Form Effects of Business Training on Financial Planning Practices (1) (2) (3) (4) (5) (6) (7) (8) Review Has Compare Has Budget Has Annual Has Annual Has Annual Has Annual Fin. Perform. Sales Sales to of Likely Profit/Loss Statement Balance Income Monthly Target Target Future Cost Statement of Cash Flow Sheet Exp Sheet Bus. Training -0.000 0.104∗ 0.061 0.100∗ 0.152∗∗ 0.052 0.069 0.070 (0.058) (0.057) (0.054) (0.053) (0.059) (0.048) (0.059) (0.059) Omitted category mean 0.561 0.529 0.284 0.226 0.439 0.213 0.374 0.516 Number of firms 327 327 327 327 327 327 327 327 Number of distinct firms 177 177 177 177 177 177 177 177 Notes: The last 2 rounds of follow-up data have been pooled for these estimations. Strata fixed effects are used in each estimation. SEs are clustered at firm level. The outcome variable in Col. (1) indicates whether the business reviews the financial performance of their business and analyze where there are areas for improvement at least monthly; Col. (2): Has a target set for sales over the next year; Col. (3): A Compares their sales achieved to their target at least monthly; Col. (4): Has a budget of the likely costs their business will have to face over the next year; Col. (5): Has an annual profit and loss statement; Col. (6): Has an annual statement of cash flow; Col. (7): Has an annual balance sheet; Col. (8): Has an annual income/expenditure sheet A.3 WALN Business Training Curriculum Modules, Sessions, Learning Objectives and Training Time Session-wise Training Time Breakdown Module I: Business Development Session Learning Objectives Training Time 1 Understanding business • To understand the meaning of business 45 minutes concepts • To explain the reasons why people, engage in business 2 Business Planning and • To understand the importance of business planning in their 7 hours Management business • To list elements of business plan 2.1 Business Plan • To exercise how to plan business start-up or expansion. 8 hours Preparation and Presentation 3 Marketing Management • To understand the importance of marketing in a business 3 hours • To plan how to use the 5 Ps of marketing in their business • To decide what promotional activities, they will use in their business 4 Financial Management and Analysis 4.1 Record keeping, financial • To understand the importance of record keeping 2 hours planning, sales and cost • To use the daily cash record plan and cash flow plan • To understand the importance of making sales and costs plans • To understand the importance of making cash flow plans 4.2 Profit and loss statement, • To prepare an income or loss statement 3 hours balance sheet and • To prepare balance sheet financial ratios • To interpret financial ratios 5 Human Resource • To understand what HRM to an organization really means 2 hours Management (HRM) • To explain the reasons why organizations, need HRM • To list guidelines which help better manage their employees 6 Negotiation skill • To understand importance of negotiation 2.5 hours • To internalize techniques of negotiation • To identify essential elements of an effective contract • To understand steps of contract negotiation 7 Stock Management • To understand the importance of stock management in a 1.5 hours business • To use stock records • To do stock records 8 Value Chain Concept • To understand concept of value chain 1.5 hours • To identify value chain actors • To internalize roles expected of value chain actors Module 1 Training Time = 31.25 hours Module II Leadership Session Skills Duration 1 Understanding • At the end of the session participants will be able to 30 minutes leadership and understand the definition and basic ideas of leadership management concepts 2 Leadership Vs • At the end of the session participants will be able to 45 minutes Management distinguish the difference between leadership and management 3 Leadership styles • Will be able to list different styles of leadership and learn 2 hours their distinct characteristics. • Will be able to recognize factors that prompt to use each leaderships style 4 Essential qualities of • At the end of the session participants will be able to list 45 minutes effective Leaders essential characteristics of effective leaders 5 Leadership and Team • will be to explain the definition of teamwork 3 hours building • Will be able to list characteristics of effective team/high performing team • Will be able to recognize the role of leaders in creating an effective team and role of followers to achieve outstanding goals 6 Leadership and decision • explain the meaning of decision making 1 hour making • Cleary comprehend types of decision making • Understand factors leaders should consider when making decisions 7 Leadership and conflict • Explain the meaning of conflict 1.5 hours management • Describe the types of conflicts • List roles of leaders in resolving conflicts 8 Personal leadership • Understand the meaning of personal development 1 hour • Define what self-esteem and its characteristics • Make an inward looking and evaluate themselves 9 Gender and leadership • Distinguish gender and sex 1.5 hours • Understand gender stereotypes • Recognize the difference between women and men leadership styles • Comprehend the challenges of women leadership Module 2 Training Time = 12 hours Module III: Communication and Networking Part 1: Communication Session Learning Objectives Training Time 1 Session 1: Meaning and • Understand the meaning and characteristics of 30 minutes Importance of communication Communication • Appreciate the importance of communication in business 2 Session 2: The Role of • Realize the role of communication in managerial activity 10 minutes Communication in a Managerial Activity 3 Session 3: Process of • Understand the process of communication 40 minutes Communication 4 Session 4: Types of • Identify the different types of communication 30 minutes Communication • Develop the skill to use different types of communications appropriately 5 Session 5: Flow of • Identify the forms and directions of communication flows 40 minutes Communication in an within an organization Organization • Familiarize themselves with the uses of different forms of communication in an organization 6 Session 6: • Identify the different styles of communication: Passive, 40 minutes Communication Styles Assertive and Aggressive • Identify their communication style 7 Session 7: Barriers to • Aware of the factors causing communication breakdown 60 minutes Effective Communication 8 Session 8: Public • Develop the skill to speak in public 40 minutes Speaking 9 Session 9: Active • Improve their listening skill 40 minutes Listening 10 Session 10: Improving • Apply principles that can improve their communication 40 minutes Communication Module 3 Part 1 Training Time = 6.17 hours Part 2: Networking Session Learning Objectives Training Time 1 Meaning Nature and • Understand what networking is 40 minutes Categories of Networks • • Identify the different categories of business networks 2 Activities Networks • Identify the key activities performed by networks 40 minutes Perform 3 Importance of Network • Realize The importance of networking for their business 40 minutes Co-operation for SMEs 4 How to Maintain Good • Understand the requirements for good networking 30 minutes Networking 5 Key Collective Actions in • Realize the requirements for collective actions in networks 30 minutes Networking Module 3 Part 2 Training Time = 3 hours Module IV: Business Mentoring/Counseling & Coaching Part 1: Mentoring/Counselling Session Learning Objectives Training Time 1 Mentoring and At the end of the session participants will be able to explain the 20 minutes mentorship meaning of mentoring and mentorship 2 Purpose of mentoring At the end of the session participants will be able to explain the 30 minutes purposes of mentoring 3 Process of At the end of the session participants will be able to explain the 30 minutes mentoring/counseling stages of mentoring 4 Types of counseling At the end of the session participants will be able to explain the 30 minutes assistance different types of mentoring counseling assistances 5 Role of mentors and At the end of the session participants will be able to list roles of 60 minutes mentees mentors and mentees 6 Key competencies of At the end of the session participants will be able to explain key 60 minutes mentor competencies of mentors Module 4 Part 1 Training Time = 3.83 hours Part 2: Coaching Session Learning Objectives Training Time 1 understanding coaching At the end of the session participants will be able to explain key 60 minutes competencies of mentors 2 Purpose of coaching At the end of the session participants will be able to explain key 60 minutes competencies of mentors Module 4 Part 2 Training Time = 2 hours Additional topics in module 4 • Mind mapping • Sensitivity/Risk Analysis • SWOT (Strength, Weakness, Opportunity, and Threat) Analysis • SCAMPER Model: o S = Substitute (change material) o C = Combine (add functions) o A = Amplify (make it bigger) o M = Minify (reduce size) o P = Put to other use (other product) o E = Eliminate (take out unnecessary parts) o R = Re-arrange (change shape or appearance)