Lesson 9/10: Firms and Markets Flashcards
Explain the motivation and background to the study on returns to capital in micro-enterprises (de Mel, 2008)
Motivation: Most of the labor force in developing countries is employed in small firms
**Research question: ** Are these “enterpreneurs by necessesity”? Or are these firms smaller than optimal due to lack of capital?
-> identification challende is selection bias
What data was used for the study on returns to capital for microenterpreneurs (de Mel, 2008)?
- Sri Lanka
- 3 districts, select 25 divisions each
- door-to-door survey to find microenterprises (1 employee, less than 1k capital stock)
- Sample covers 408 firms (203 in retail, 205 in manufacturing
- surveyes quarterly for 2 years (9 waves)
- data comes exclusively from own survey (self-reported profits, entrepreneurs traits, including risk aversion)
Explain the RCT design in the paper on returns to capital for microenterpreneurs (de Mel, 2008)
Goal: measure the returns to capital in microenterprises
* 228 out of 408 firms randomly assigned to receive:
* T1: 100$ cash
* T2: 200$ cash
* T3: 100$ in kind
* T4: 200$ in kind
where 100$ = 3 moths median profit
-> grants frames as reward for survey
What are the findings of the paper investigating returns to capital in microenterprises (de Mel, 2008)?
Basic treatment effects:
Model: Y 0 a + Σ ß Treatment + Σ δ + γ + ε
where
δ: survey wave FE
γ: captures individual specific propensities
- all T firms used the grants to expand existing business
Findings:
* treatment increases capital, profits, hours worked by owner
* using different adjustments for owner’s additional work effor, estimate returns of 4.6% - 5.3% per month, i.e. 55% -63% per year
* effects are larger for owners with these characteristics: poor, better educated, more skilled, male, employ less household labor
* no difference by risk aversion
What were the required identifying assumptions in the paper on returns to capital in microenterprises (de Mel, 2008)?
- SUTVA -> no spillovers
- Findings: spillovers only in the bamboo industry (due to governmental restrictions on bamboo harvesting)
- no (systematic) attrition
- Exclusion restriction: treatment affects firm profits only through its impact on capital
- Findings: effects on owner’s labor violates this -> authors adjust profits for effect of owners time
Explain the identifying assumption around the LATE to equal the average marginal return to capital in de Mel, 2008.
- if marginal return to capital is the same for all firms, then after the adjustment fro own labor hourse, the IV estimator will be the average marginal return to capital
- if the conditions do not hold and we assume that the change of capital stock due to treatment is non-negative, then the IV provides a LATE which is a weighted average of the marginal returns to capital, with the marginal return to each firm weighted by how much that firm’s capital stock responds to the treatment -> no suggestive evidence for this
Explain the wide dispersion in the results on returns to capital for microenterprises (de Mel, 2008).
- half of the female owners have characteristics suggesting negative returns and 60% have returns lower than market interest rates of 12 -18%
- in male-owned enterprises, just over 20% of the enterprises have predicted returns that are lower than market interest rates
-> this may explain why so few of the enterprises borrow from formal lenders
-> household liquidity and ability as useful screening devices to identify microenterprises where investments are likely to be most profitable
High variance in returns may explain why these firms are too risky for formal lenders
Explain adverse selection and give an example.
- if lender cannot distinguish between risky and safe type of borrower -> the higher the fraction of risky type in the market, the higher the interest rate, the higher fraction of risky type etc.
Explain limited liability and give an example.
- financing decisons also depend on how much of your own wealth is claimable (liable)
Explain the research and issues around microfinance.
- in theory, microfinance may mitigate credit market failures by harnesssing local information & lowering monitoring and enforcement costs
Research - it has no (or very limited effect in practice
- Meta-study of 7 RCTs of microfinance (Meager, 2019, 2022) -> no transformative impact on household business outcomes, except for those at the top of the profit distribution // reasonable external validity of these impact evaluations
- some investigations of possible welfare-enhancing tweaks to the standard microfinance contracts
**Issues: **
* small firms may require more capital than the size of the typical microfinance loan would fund in order to grow (e.g. rental market for production machines between small informal manufacturing firms in Uganda)
* Recent studies generate experimental vairaion in the size of laons to test this hypothesis:
-> Bari et al. (2022): find large and persistent effects from offering microcredit clients asset financing for up to 4x ususal borrowing limit in Pakistan
-> In Egypt, Bryan et al (2024) find that larger loans generate small average impacts, but with larger dispersion in treatment effects on profits, in line with ML predictions from psychometric data
Give some intuition on management practices and the connection to access to inputs.
Data: world management survey
* survey for broad practices spanning operations/monitoring, target setting, and people/ incentive management practices
* based on semi-structured interviews with middle managers
* examples of good practices include: having reasonable number of KPI and keep track of them regularly, having a system in place to identify which workers are perfomring wel and which ones need help to improve
Insights
* Firms in poorter countries tend to have poorer management
* management quality varies across and within countries -> even in lower ranking countries, there are firms that adopt practices well
**Possible explanations: **
* ownership structure: higher prevalence of family firms, which tend to be less well managed, in poorer countries
* product market competition is lower in poorter countries
* worker skills are lowers (true for both managers and employees)
* labor market regulations matter for the adoption of incentive management practives
* informational constraints??
Explain the paper “Does management matter?” (Bloom et al. 2013)
Research about India
Identification challenge:correlations between management score and firm performance may be misleading
Solution: a (very expensive) randomized experiment
* focus on one industry (woven cotton) to implement common set of practices and measure a common set of outcomes
* 17 indinan texttile firms with 28 plants in total
* Management consulting provided by major international firm (intital diagnostic phase received by both T & C plants, 4 months of free consutling (T only)
**Findings: **
* improvements in management practices led to a 17% improvement in productivity in the first year
* quality defect were cut by 50%, inventories by 40%
* firms profits increased by about 325k$ per plant per year (market cost of the intervention 200k$)
* improved the ability of owners to expand their firms (new production plants openend within 3 years)
* Follow-up study several years later (Bloom et al, 2020): half of the practices originally adopted were still in use, but had spread across untreated plants within the same firm
What does the paper on management practices suggest?
- results point to informational battiers as key constraint to adoption (either because they were considered not to improve profits or because they were not known about
- consistent with the fact that most managers over-estimate own management quality
-
But competition did not force badly run firms to exit, why?
-> constraints on growth derived from limited managerial span of control ( hinder well-run firms to take over the market)
-> family members did not trust non-family members to do a good managerial job
Explain the role of market demand.
- much of the literture emphasizes supply-side constraints (assuming demand will follow once supply improves)
Goldberg & Reed (2023): * demand-side constraint can be binding - especially in low-income countries with small markets - structural transformation requires firms adopting increasing returns to scale (IRS) technologies -> only feasible if aggregate demand is large enough to justify the FC
- demand size can come from: a sizeable domestic middle class, international trade access
-> some countries may be “ too small to develop” without trade or domestic redistribution to expand market demand
Explain the importance of output markets for firm growth in the paper by Hardy & Kagy (2020) and elaborate on the paper in detail.
- study the role of demand constraints in the gender profit gap documented in the Ghanaian garment sector
Data: firm census and market research survey
1) census of all garment-manufacturing firms owener in Hohoe, mid-sized district capital in Ghana (291 female-owned, 92 male owned firms)
2) Market research survey: female and make customers order about the same quantities, gender segregation in demand: women tend to buy from female-owned firms, men from male-owned firms
-> point to a gender gap in the market size to firm ration, suggesting a demand consraint for female-owned firms
**Findings: **
* female owned companies report not eough customers as a barrier to business success more than male-owned firms
Experimental Demand Boost:
* experimentally increased demand for both female-owned and male-owned firms’ putput with orders of 0 - 10 garments
* Key findings: female-owned firms absorb all demand without displacement of existing orders -> suggests that they are not operating at capacity, not the case for male-owned firms
-> overcrowding in the female-owned garment manufacturing sector contributes to the gender profit gap
Explain the importance of output markets for firm growth in the paper by Hjort et al (2021).
- do informational barriers prevent access to high-paying buyers?
- offers a random subset of medium.sized Liberian firms vouchers for a week-long training on how to sell goods and services to governments, corporations, and other large buyers
- find significant impacts on performance, largerly concentraded in one quartile of initially disadvantaged firms
- three years post-training, these firms continue to win desireable contracts, are more likely to operatie and more more workers
Explain the paper on Market Integration by Jensen & Miller (2018)
Follow up paper to Jensen (2007)
**Original paper: **
* studies the effects of cell phone coverage on price dispersion, waste and consumer + producer surplus in fishing markets in Kerala, India
Hypothesis: inefficiencies could arise becaue localised markets and limited communication infrastructure
* information is central to economic theory: law of one price (price of a good should not differ between any two markets by more than the transport cost between them) -> but information is in practice limited
-> this paper exploits region-by-region roll-out of cell phone coverage in Kerala, India to evaluate the effects of improved market information
**Findings: **
* mobile phones reduce price dispersion
**Follow-up Paper: ** Market Integration, Demand and the Growth of Firms
A set of stylized fact motivates this paper:
1) Firms in developing countries generally start small and stay small
2) these firms also tend to have low productivity
3) Significant productivity dispersion across firms suggests that competition is not driving out the least productive firms
Research Question: How does expanding a firm’s customer base beyond its local market affect its survival, size and market share?
Hypothesis: lack of market integration from information frictions, firms are often unable to sell beyond a fairly local market because it is difficult for consumers to learn about the existence and quality of different firms’ output.
-> limited size of their potential customer based limits goods firms’ ability to grow and prevents them from exploiting economies of scale
-> lack of competition also allows less productive firms to survive
What is the methodology & identification strategy in Jensen & Miller?
Methodology:
* Least squares regressions
* Model: Y = a + a1 Phone + a2 Quality + a3 Phone x Quality + ε
**Identification Strategy: **
* mobile phones provide exogenous shock to market integration
* Assumption: had it not been for the introduction of mobile phones, there would have been no differential change in these outcomes across builders
What are the findings and results from Jensen & Miller?
1) Changes in fishermen’s behavior and information (less percent sell fish in their own village, start buying boats in other villages and less errors in estimating boat life-span)
2) Changes in exit, market share and firm size (reduction in total number of firms, more bad firms exit the market)
3) Changes in output and quality
4) Changes in cost
5) Economies of scale and returns to labor specialization -> within-firms efficiency gains (same number of boats with 25% fewer labor hours and 37% less capital // firm growth associated with significant labor specialization // decrease in average production cost)
6) Gains for consumers (fishermen) -> even though the raw sales price of boats increased slightly, the lifespan of a boat purchased increased by 1.35 years, so price per boat-life purchased declined by 23%
What are the conclusions from the paper by Jensen & Miller?
- factors such as imperfect information that limit the ability of businesses to get customers outside of their local area are an important constraint on firm growth
- firms were able to grow withou greater access ro credit, improved infrastructure or changes in any other factor
- credit cosntraints or managerial skills may become more important in out setting beyond an early staeg growth, since although firms grow considerable, they still remain fairly small
- clear role of limited effective market size as one potential constraint on growth and productivity
Explain the evidence on export markets and expanding market size (Atkin et al. (2017))
Motivation: stylized fact that exporting enterprises are more productive than enterprises that only sell domestically, is this driven by (only) selection effects? Or could encouraging firms to export raise firm performance?
-> **Need exogenous shock to exporting to assess causal effect **
**Setting & Experimental Design: **
* 305 Egyptian carpet producers
* NGO and local distributor marekt handmade carpets to US and Europe importers
* Random subset of firms are offered the opportunity to fill these orders
**Findings: **
* Treatment Effects: around 0.20 TE (ITT)
* Mechanism: learning-by-exporting
Evidence:
1) both quality and productivity rise afer adjusting fro product specifications
2) Quality lab -> higher quality in production
2) Documented learning curves for both
3) correspondence drawn between foreign buyers adn the intermediary -> knowledge flows
4) rule out investment explanations by showing that treatment firms make not monetary or time investments in upgrading
