Finance & PSD Impact SEPTEMBER 2020 The Lessons from DECFP Impact Evaluations ISSUE 57 Our latest note examines how sensitive the demand for business training is to price, and whether charging a higher price causes firm owners to exert more effort in training. How should business training be priced? A demand experiment in Jamaica Alessandro Maffioli, David McKenzie and Diego Ubfal At least one billion dollars is spent annually firm. The firms were microenterprises in a wide training at least 4 to 5 million potential and range of industries, with median monthly profits existing entrepreneurs in developing countries. of $30-60 in our two samples, so the full cost of In the majority of cases, including in many the course would be 3-5 months of profits. World Bank projects, this training is given by We began by inviting firms to attend governments to firms for free. demonstration sessions, where they could learn Free training is often justified on the grounds about the course, get examples of its content, that firms are too poor or credit-constrained to and hear results from other entrepreneurs that afford the costs of training, and that poorly managed firms may underestimate the benefits had taken part previously. At the end of the of improving their business practices. session, firms were then offered the opportunity However, these programs use public money to purchase the course. to provide private benefits for the participating The first experiment uses the Becker- firms. Charging a positive price may improve deGroot-Marschak (BDM) mechanism with 457 the financial sustainability of such programs, entrepreneurs. This asks firms to state their and enable governments to train more firms with maximum willingness to pay for the course. A a given budget. Higher prices may also help price is then randomly drawn, and if this price is target training towards the firms that expect to below the maximum willingness to pay, the benefit most from it. Finally, a sunk cost effect entrepreneur buys the course at the drawn price. may operate, whereby paying for training causes For example, if the entrepreneur is willing to pay firm owners to value it more, and exert more $50, they get the course for $30 if a price of $30 effort in learning from training. These arguments are all theoretically is drawn, but are not allowed to purchase the plausible, but what is missing is empirical course if a price of $60 is drawn. evidence on how sensitive the demand for The second approach uses the simpler training is to price, and on whether this sunk cost take-it-or-leave-it (TIOLI) approach with 374 effect operates in practice. firms. Here one of four prices (free, $37, $75, $112) was drawn, and the firm could decide 1. A Demand Experiment in Jamaica whether or not to buy the course at this price. We conducted two pricing experiments to In both experiments we also randomized measure these effects in partnership with the whether entrepreneurs had to pay for the whole Jamaica Business Development Corporation. In course at once, or whether they could receive both cases the training course offered was a 40- credit and pay in three installments. hour course that had modules on improving business practices, and on personal initiative. 2. Results The cost of providing the course was $150 per Charging higher prices reduces demand Figure 1 shows how the demand for business Higher prices change who is trained: higher training varies with price in the two prices screen out poorer, older, and more risk- experiments. The full cost of the training was averse entrepreneurs with smaller businesses. J$20,000 ($150). We see that the share of However, higher prices do not differentially entrepreneurs who purchase the course when it affect female business owners. is offered at its full cost is very low. However, Higher prices both select entrepreneurs who are demand is inelastic, so that doubling the price more likely to attend training, and have a sunk cuts demand by less than one half. cost effect: entrepreneurs with a higher willingness to pay are more likely to attend Figure 1: Demand Curve for Business when selected for training; and conditional on Training willingness to pay, those who are randomly assigned higher prices and pay for training go to more sessions. 3. Policy Recommendations Our results suggest the optimal price for business training is unlikely to be either free or full cost. Charging for training does screen out those firm owners who are less likely to attend, selects those who expect to benefit more from it, and induces more effort in attending training. But a lot of these effects occur from charging only one-quarter of the full cost, and charging As a result, fewer people are trained at higher close to or at full recovery cost ends up prices: In our BDM sample, 76% of screening out many business owners and entrepreneurs attend at least one training class increasing recruiting costs. when it is free, 65% when they have to pay a Finally, policy efficiency would suggest nominal fee of 5%, 43% when charged one- targeting training at those firms that will benefit quarter of the cost, 29% when charged half the most from it. A key question is then whether the cost, and only 11% when charged the full cost. entrepreneurs who have higher willingness to In the TIOLI sample, which consists of pay actually benefit more from training. We somewhat wealthier entrepreneurs with larger were unable to measure this with the sample firms, 90% attend at least one class when it is sizes in our study, and this is an important topic free, and demand falls to 48% when charged half for future work to evaluate. the cost, and 37% when charged three-quarters of the cost. In both cases, offering the opportunity to pay in three installments did not affect demand. For further reading see: Maffioli, A., McKenzie, D. and Ubfal, D. “Estimating the Demand for Business Training: Evidence from Jamaica”, World Bank Policy Research Working Paper 9415, September 2020. Recent impact notes are available on our website: https://www.worldbank.org/en/research/brief/finance-and- private-sector-impact-evaluation-policy-notes