Scribd
Upload
7 views24 pages

Slides 20 CapitalMisallocation1

Uploaded by

Kalkidan Getachew
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
7 views24 pages

Slides 20 CapitalMisallocation1

Uploaded by

Kalkidan Getachew
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 24

Factor Misallocation

Capital Misallocation: Micro Evidence


Taking stock
• Still trying to understand what drives TFP
• Key to understand why different countries produce vastly different
amounts controlling for factor endowments
• In the last lectures we saw how this can be explained by things that
make factors more productive
• National institutions, management
• Next, we study how factors are allocated across firms, and whether
misallocation can explain differences in TFP

NYUAD - Economic Growth - ECON UH 3030 2


Misallocation?
• With perfect markets, factors flow to their most productive use until
returns are equalized
• This guarantees that output is maximized for a given quantity of
factors
• Should not be possible to move factors from one firm to another and
increase output
• With market imperfections, this might not occur
• Thus, two countries with the same factors might have different
outputs if the allocation differs

3
Maximizing output given capital stock 𝐾
• Assume you want to allocate 𝐾 between two firms to maximize total
output value:
max 𝑝𝑖𝑓𝑖(𝑥𝐾) + 𝑝𝑗𝑓𝑗((1 − 𝑥)𝐾)
!

• First-order condition:
𝑝" 𝑓"# (𝑥 ∗ 𝐾) − 𝑝% 𝑓%# ((1 − 𝑥 ∗ )𝐾) = 0
• Output is maximized when K is allocated so that the two firms have
the same marginal revenue product
• What if 𝑥 < 𝑥 ∗ ?

NYUAD - Economic Growth - ECON UH 3030 4


Efficient allocation through markets
• In perfect markets all firms face same price of capital r
• They choose their level of capital 𝑘 to maximize 𝑝𝑖𝑓𝑖(𝑘) − 𝑟𝑘
• Thus, in equilibrium 𝑝𝑖𝑓’𝑖(𝑘) = 𝑟 for all 𝑖
• Hence, 𝑝𝑖𝑓’𝑖(𝑘) = 𝑝𝑗𝑓’𝑗(𝑘) for any i-j pair
• Firms have the same revenue productivity à output is maximized
given K
• Note: physical productivity 𝑓’𝑖(𝑘) might differ but revenue
productivity 𝑝𝑖𝑓’𝑖(𝑘) is the same

NYUAD - Economic Growth - ECON UH 3030 5


Misallocation of capital
• We say there is misallocation if 𝑝𝑖𝑓’𝑖(𝑘𝑖) ≠ 𝑝𝑗𝑓’𝑗(𝑘𝑗)
• Say 𝑝𝑖𝑓’𝑖(𝑘𝑖) > 𝑝𝑗𝑓’𝑗(𝑘𝑗) à firm j has “too much” 𝑘
• This reduces total output
• Different extent of misallocation can explain differences in TFP
• But why would there be misallocation?

NYUAD - Economic Growth - ECON UH 3030 6


Market imperfections and misallocation
• Imperfections might lead to different firms facing different 𝑟
• E.g., asymmetric information in credit markets creates need for
collateral à firms with more collateral face lower 𝑟
• Say 𝑖 has more collateral than 𝑗, then
𝑝𝑖𝑓’𝑖(𝑘𝑖) = 𝑟𝑖 < 𝑟𝑗 = 𝑝𝑗𝑓’𝑗(𝑘𝑗)

• Larger (𝑟𝑖 − 𝑟𝑗) wedge à more severe misallocation

NYUAD - Economic Growth - ECON UH 3030 7


Sources of misallocation
• Imperfect information
• Policy
• Taxes
• Labor regulation
• ..
• Political connections
• Externalities

NYUAD - Economic Growth - ECON UH 3030 8


How much misallocation?
• Theoretical implications are fairly clear:
• Misallocation can reduce TFP
• But how important is this in practice?
• Next: two methods to find an answer:
1. Experimentally vary capital of firms to estimate returns (DeMel et al. 2008)
• If returns are different across firms -> evidence of misallocation
2. Calibrate theoretical model (Hsieh and Klenow 2009)

NYUAD - Economic Growth - ECON UH 3030 9


Returns to Capital in Microenterprises:
Evidence from a Field Experiment
DeMel, S., D. McKenzie, and C. Woodruff, QJE 2008

NYUAD - Economic Growth - ECON UH 3030 10


Questions
• A large share of the labor force in developing countries is employed in
small firms (1-5 people)
• Are these firms smaller than optimal? And if so, what’s stopping them
from growing?

NYUAD - Economic Growth - ECON UH 3030 11


This paper
• Implements a field experiment to measure the returns to capital in
microenterprises
• Allocate cash or in-kind grants to randomly selected microenterprises
in Sri Lanka
• Measure returns as the difference in profit between treatment and
control firms
• Is there heterogeneity in returns to capital?
• Are the returns equal to the interest rate?

NYUAD - Economic Growth - ECON UH 3030 12


Study sample
• Three districts, select 25 divisions each
• Door-to-door survey to find microenterprises
• 1 employee (the owner)
• K<$1000
• Sample covers 408 firms
• 203 in retail (grocery stores)
• 205 in manufacturing/services
• surveyed quarterly for 2 years (9 waves)

NYUAD - Economic Growth - ECON UH 3030 13


Treatments
• 228 out of 408 firms randomly chosen to receive
• T1: $100 cash
• T2: $200 cash
• T3: $100 in kind (inventories/materials/machinery)
• T4: $200 in kind (inventories/materials/machinery)
• $100= 3 months median profit
• Framed as random prize drawing to compensate for survey
• All treatment firms used the grants to expand existing business

NYUAD - Economic Growth - ECON UH 3030 14


Data
• Exclusively from own surveys
• Self reported profits
• Entrepreneur traits
• Gender, age, education
• Risk aversion, numeracy

NYUAD - Economic Growth - ECON UH 3030 15


Balance table

Should we
worry that
household
assets are
significantly
higher in
control
group?

NYUAD - Economic Growth - ECON UH 3030 16


Estimation

• 𝑌"& : outcome of firm 𝑖 in wave 𝑡


• 𝑔 indicates the treatment arm (1-4)
• 𝛿& : wave fixed effects (what does that control for?)
• 𝜆" : firm fixed effects (what does that control for?)
• Standard errors clustered at firm level

NYUAD - Economic Growth - ECON UH 3030 17


Treatment increases capital and profits

NYUAD - Economic Growth - ECON UH 3030 18


Treatment also increases labor - why?

NYUAD - Economic Growth - ECON UH 3030 19


Estimating returns to capital

• Instrument capital with treatment


• Exclusion restriction: treatment affects profits through 𝐾 alone
• Effect on 𝐿 violates this à adjust profits for effect on owners’ time
• We saw treatment increases labor
• Labor also affects profits
• Adjust to ensure we only capture marginal return to capital net
of effect of increased labor
NYUAD - Economic Growth - ECON UH 3030 20
NYUAD - Economic Growth - ECON UH 3030 21
Estimated returns
• 𝑟 = 5.85% per month
• 𝑟 = 4.59% with labor adjustment
• Return higher for
• poorer
• fewer workers in the HH
• better educated
• more skilled
• men
• No difference by risk aversion

NYUAD - Economic Growth - ECON UH 3030 22


Lessons
• Average return of at least 55% per year
• 4.59% monthly returns * 12=55 (assume no reinvestment, lower bound)
• Prevailing interest rate: 12-18%
• Hence, capital is misallocated
• Why?
• High variance in returns can explain why these firms are too risky for formal
lenders (credit market failure)
• But why don’t they save their way out?

NYUAD - Economic Growth - ECON UH 3030 23


Take aways
• RCTs have the benefit of clean identification
• But there is a trade-off between precision and generality
• DeMel et al. (2008) give us a clean causal estimate that suggests a
large effect of misallocation on TFP in a specific setting
• But they cannot quantify its aggregate impact at the country level
• Next lecture, we will see a paper that aims to do exactly this

NYUAD - Economic Growth - ECON UH 3030 24

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy