20 01 16 6 A Af fr ri ic ca an n E Ec co on no om - - PowerPoint PPT Presentation

SLIDE 1

2 20 01 16 6 A Af fr ri ic ca an n E Ec co

• n

no

• m

mi ic c C Co

• n

nf fe er re en nc ce e C CR

RE ED DI IT T

C CO

ON NS ST TR RA AI IN NT TS S

A AN

ND D

F FA

AR RM M

P PR

RO OD DU UC CT TI IV VI IT TY Y:

: M

MI IC CR RO O-

• L

LE EV VE EL L

E EV

VI ID DE EN NC CE E

F FR

RO OM M

S SM

MA AL LL LH HO OL LD DE ER R

F FA

AR RM ME ER RS S

I IN

N

E ET

TH HI IO OP PI IA A

Adamon N N. . Mukas asa, , Anthony M. Simpasa, and Adeleke O. Salami

Development Research Department, African Development Bank, Abidjan, Cote d’Ivoire

5-7 December 2 2016, A Abuja, , Nigeria

SLIDE 2

Outline

1. 1. Introduction  Background  Motivation and objectives of the study 2. 2. Methodology  Model of farmers’ access to and demand for credit  Credit constraints and farm productivity  Productivity loss due to credit constraints 3. 3. Data an and d descriptive statistics 4. 4. Results 5. 5. Conclusio ion a and r recommendations

SLIDE 3

Introductio ion (Background)

 Positive correlation between access to credit and targeted outcomes (Feder et al, 1990; Sial and Carter, 1996; Carter and Olinto, 2003; Foltz, 2004):

• Acquisition on production inputs (short term) and realisation of

profitable investments (long term)

• Transition from subsistence farming to agri-business
• Poverty reduction

 Imperfe fections of credit markets are the norm (Stiglitz and Weiss, 1981; Carter, 1989; Feder et al, 1990):

• Asymmetric information (adverse selection and moral hazard) in

the credit markets

• Misallocation of resources and a sub-optimal use of inputs
• Lower income, volatile welfare and food insecurity (Jappelli, 1990;

Petrick, 2004; Ali et al, 2014)

SLIDE 4

Introductio ion (Background)…

 In Eth thio iopia, credit market imperfections are more acute in th the agric icult ltural l sector:

• Small

ll-scale le farmers (crop production is dominated by subsistence farm households and more than 90% 90% of the farms with at most 10 acres)

• Relia

iance on own resources for input and consumption purchases, labor hiring, investments,…

• Vulnerabil

ilit ity to vagarie ies of nature (droughts, floods,…) and to market t in insta tabil ilit itie ies (price volatility, high transaction costs,…)  Improvements ts in financial inclusion but still in insuffic icie ient:

• Agriculture receives less than 10 percent of the banks’ lending
• 22%

22% of adult Ethiopians hold an account at a financial institution in 2014 vs 29% 29% in SSA, 45% 45% in South Asia, 51% in Latin America, and 69% in East Asia

• Only 3 bank branches per 100,000 inhabitants in 2014 vs 4 branches in

SSA countries and 14 14 in the world.

SLIDE 5

Introductio ion (Motivation)

 How limited access to credit (or lack thereof) affects farm productivity?

• Are cr

credit constrained and unconstrained farmers intrinsically different?

• To what extent credit constraints affect farm p

productivity in Ethiopia?

• What is the potential productivity gain from removing credit

constraints in Ethiopia?  Our contributions to the understanding of the potential linkages between credit constraints and farm productivity:

• Analysis of the nature and det

eterminants of credit constraints in Ethiopia’s agricultural sector (supply- and demand-side factors)

• Use of econometric techniques to disentangle the differential e

l effects o

• f

f various t types of c credit c constraints on farm productivity

• Evaluation of the potential productivity lo

loss due to credit constraints (or potential gains from removing them)

SLIDE 6

Methodology

 Access to and demand for credit

• Simultaneous estimation

(latent notional demand for credit)

( 0) (observable demand for credit)

(latent notional access to credit)

( 0) (observable access to credit)

• Estimation

using bivariate probit model with partial

• bservability:

( , ) ( , ) (1 ) ( , )

SLIDE 7

Methodology…

 Credit constraints and farm productivity ty

• Issu

sues: unobserved heterogeneity and endogenous sample selection

• Way out: endogenous regime switching regression model (Freeman et al, 1998;

Lokshin and Sajaia, 2004; Ali and Deininger, 2012) using FIML method: 1

1

 Productiv ivity l loss due to credit constraints 1

Selection equation Productivity equations

SLIDE 8

Data and descriptive statistics

 2011-2012 Ethiopia Rural Socioeconomic Survey (ERSS) and 2013- 2014 Ethiopia Socioeconomic Survey (ES ESS) as part of the Living Standards Measurement Study – Integrated Surveys on Agriculture (LSMS-ISA)  Three r rounds p per s survey:

• The first round in September and October (2011 for ERSS and

2013 for ESS) for information on households’ post-planting activities.

• The second round in November-December (2011 for ERSS and

2013 for ESS) for livestock activities.

• The last round in January-March 2012 for ERSS and in February-

April 2014 for ESS with post-harvest information  Balanced sample of 2, 2,654 h households engaged in agricultural activities

SLIDE 9

Data and descriptive statistics…

 Defi finition o

• f

f credit c constraint s status: Direct e elicitation a approach

NON Why not? YES Received total amount applied for? Unconstrained Applied for a loan in the past 12 months? NON YES

• Loan rejected
• Partially approved

Quantity constrained

• No need, have enough resources

Unconstrained

• Interest rates too high

Price constrained

• Inadequate collateral
• Absence of lender nearby
• Do not know where to apply
• No bank account
• Too much trouble
• Do not like to be in debt
• Believed would be refused
• Fear not be able to pay

Transaction cost constrained Risk constrained

SLIDE 10

Descriptive statistics…

Pooled sample 2011-12 2013-14 Number Relative frequency Number Relative frequency Number Relative frequency Unconstrained 1,771 33.36 520 19.59 1,251 47.14 Constrained 3,537 66.64 2,134 80.41 1,403 52.86 Quantity- constrained 264 7.46 145 6.79 119 8.48 Price- constrained 238 6.73 147 6.89 91 6.49 Risk-constrained 2,546 71.98 1,555 72.8 .87 991 70.6 .63 Transaction costs-constrained 507 14.33 294 13.78 213 15.18 Observations 5,308 100 2,654 100 2,654 100

SLIDE 11

Descriptive statistics…

Description All panel households Credit constrained No Yes Household characteristics Age of the head (in years) 45.66 (15.18) 45.22 (14.31) 45.88 (15.59) Education of the head (in years) 1.70 (3.04) 1.73 (2.97) 1.69 (3.07) Female-headed household (in %) 20.43 18.74 (39.03) 21.28(40.94)** Household size 5.28(2.23) 5.40 (2.21) 5.22 (2.23)*** Monthly consumption ( ETB) 603.21 (1,808.46) 580.02 (1,371.43) 614.82 (1,991.66) Access to extension services (%) 15.79 (36.47) 12.76 (33.38) 17.30 (37.83)*** Farm characteristics Crop output (in ETB) 9,736.68 (81,854.51) 14,879.73 (139,201.5) 7,146.51 (18,343.46)*** Land size (in acres) 3.82 (17.38) 4.26 (13.25) 3.60 (19.11) Land productivity (ETB/acre) 6,783.94 (63,496.29) 10,906.07 (106,647.5) 4,719.95 (18,575.11)** Family labor (in hours) 1,089.86 (1,976.07) 1,147.31 (1,442.52) 1,061.09 (2,194.65)*

SLIDE 12

Descriptive statistics…

All ll p panel h household lds Credit c constrained No Yes Hired labor ( person-days) 56.67 (390.57) 82.38 (614.04) 43.80 (199.28)*** Chemical fertilizer use ( kg/acre) 8.97 (23.65) 10.17 (21.72) 8.37 (24.53)** % of plots with certificate 36.20 (43.58) 39.06 (43.38) 34.77 (43.62)*** Land slope (% of flatted plots) 54.95 (36.67) 58.27 (35.66) 53.29 (37.06)*** Distance to nearest main road (km) 16.70 (19.81) 16.18 (17.93) 16.96 (20.68) Distance to main market (km) 67.27 (48.97) 65.03 (46.53) 68.39 (50.12)** Distance to main population center (km) 37.57 (26.52) 37.53 (26.34) 37.60 (26.62) Distance to nearest commercial bank (km) 23.81 (26.45) 22.50 (27.23) 24.46 (26.03)** Distance to nearest microfinance institution (km) 13.22 (19.01) 14.00 (18.77) 12.83 (19.12)** Tropical Livestock Units 8.74 (10.83) 8.83 (10.68) 8.70 (10.91) Asset index 0.17 (3.61) 0.06 (2.05) 0.23 (4.18)* Observations 5,308 1,771 3,537

SLIDE 13

Results (1)

 Determinants o

• f

f access t to a and d demand fo for credit i in Ethiopia

Bivariate probit with partial observability Access Demand Farm size (acre) 0.025 (0.014)** 0.035 (0.016)** Household size (number) 0.07 (0.019)*** 0.026 (0.006)*** Adult rate (%) 0.458 (0.133)***

• 0.157 (0.168)

Female (female=1)

• 0.089 (0.090)

0.113 (0.112) Age (years) 0.000 (0.003)

• 0.017 (0.003)

Age squared 0.001 (0.000) 0.002 (0.002) Years of schooling 0.006 (0.010)

• 0.016 (0.013)

Share of titled lands (%) 0.131 (0.076)* 0.005 (0.102) Objective of credit (Input purchases=1) 2.497 (0.144)*** 2.203 (0.141)*** Tropical Livestock Units (number) 0.010 (0.005)***

• 0.003 (0.007)

Asset index 0.042 (0.053)

• 0.007(0.025)

Rural (Rural=1) 0.063 (0.163)

• 0.330 (0.245)

Commercial bank in the village (Yes=1) 0.117 (0.278) Distance to commercial bank (km) 0.002 (0.005) Microfinance institution in the village (Yes=1) 0.464 (0.163)*** Distance to microfinance institution (km) 0.008 (0.006) Distance road (km)

• 0.007 (0.007)

Distance market (km) 0.002 (0.001) Distance population center (km)

• 0.00 (0.004)**

Survey wave 2 (Panel 2=1) 0.024 (0.015) 0.016 (0.009)* Constant

• 1.38 (0.238)***

2.932 (2.617)

SLIDE 14

Results (2)

 Switchin ing e estimation: d determin inants o

• f being

credit c constrained

All constrained (Model I) quantity- constrained (Model II) Price- constrained (Model III) Risk- Constrained (Model IV) Transaction costs- constrained (Model V)

Female 0.074 (0.053) 0.078 (0.206)

• 0.172

(0.097)* 0.082 (0.048)* 0.003 (0.070) Age

• 0.018

(0.008)**

• 0.001

(0.014) 0.018 (0.015)

• 0.018

(0.010)* 0.005 (0.011) Age squared 0.000 (0.000)**

• 0.000

(0.000)

• 0.001

(0.000) 0.000 (0.000)**

• 0.000

(0.000) Years of schooling

• 0.005

(0.006)

• 0.001

(0.001)

• 0.005

(0.012)

• 0.013

(0.006)**

• 0.012

(0.009) Farm size

• 0.005

(0.021)

• 0.091

(0.049)* 0.043 (0.031)

• 0.031

(0.045)

• 0.007

(0.033) Objective of credit

• 0.605

(0.059)***

• 0.364

(0.060)***

• 0.095

(0.013)***

• 0.240

(0.349)

• 0.099

(0.258) Rural

• 0.252

(0.128)**

• 0.343

(0.283) 0.205 (0.281) 0.024 (0.129) 0.031 (0.153)** Commercial bank in the village

• 0.469

(0.120)***

• 0.041

(0.223) 0.097 (0.253)

• 0.106

(0.147)

• 0.061

(0.134) Distance to commercial bank 0.001 (0.001) 0.001 (0.001) 0.000 (0.001) 0.002 (0.001)* 0.002 (0.001)**

• Micro. institution

in the village 0.181 (0.051)***

• 0.072

(0.081) 0.143 (0.178) 0.069 (0.125)

• 0.104

(0.067) Distance to micro. institution 0.001 (0.001)

• 0.001

(0.002) 0.000 (0.003) 0.003 (0.003) 0.005 (0.002)*** Distance market 0.001 (0.001)*** 0.000 (0.002) 0.000 (0.002) 0.000 (0.003) 0.000 (0.001) Distance population center

• 0.001

(0.001)

• 0.001

(0.004)

• 0.001

(0.002)

• 0.001

(0.004) 0.003 (0.001)** Survey wave 2

• 0.262

(0.117)** 0.570 (0.387) 0.347 (0.298)

• 0.094

(0.062) 0.084 (0.161) Constant 0.573 (1.147)

• 7.756

(1.550)*** 1.283 (1.774) 1.392 (1.259)

• 3.761

(1.403)***

SLIDE 15

Results (3)

 Endogenous regime switching r regression m model – Second stage ( (All c constraints)

Constrained Unconstrained Household size

• 0.065

(0.031)**

• 0.037

(0.034) Adult rate

• 0.367

(0.199)*

• 0.582

(0.214)*** Farm size

• 0.595

(0.058)***

• 0.478

(0.065)*** Share of irrigated plots 3.522 (0.425)*** 3.062 (0.508)*** Family labor 0.984 (0.048)*** 1.153 (0.056)*** Hired labor 0.238 (0.036)*** 0.205 (0.036)*** Use of improved seeds 0.502 (0.142)*** 0.376 (0.117)*** Use of fertilizer 0.709 (0.194)*** 0.522 (0.223)*** Use of agro-chemicals 0.435 (0.113)*** 0.057 (0.111) Use of manure 0.363 (0.149)** 0.146 (0.137) Average rainfall 2.453 (0.445)*** 1.828 (0.492)*** Survey wave 2 1.430 (0.285)*** 2.011 (0.458)*** Mills ratio 0.559 (0.004)*** 1.167 (0.005)*** Constant

• 20.449

(2.716)***

• 9.299

(3.518)***

SLIDE 16

Results (3)

 Endogenous r regime s swit itching r regressi sion m model l – Second s stage b by t type o

• f

f credit c constraint nts

Quantity constrained only Price constrained only Risk constrained only Transaction cost constrained

• nly

Constrained Unconstrained Constrained Unconstrained Constrained Unconstrained Constrained Unconstrained Household size 0.027 (0.096)

• 0.053

(0.025)**

• 0.199

(0.205)

• 0.067

(0.025)

• 0.008

(0.041)

• 0.044

(0.032)

• 0.071

(0.087)

• 0.024

(0.027) Adult rate

• 0.347

(0.649)

• 0.389

(0.159)**

• 2.002

(0.957)**

• 0.299

(0.151)** 0.619 (0.781)

• 0.498

(0.197)**

• 0.329

(0.414)

• 0.209

(0.173) Years

• f

schooling 0.016 (0.047) 0.012 (0.016) 0.016 (0.082) 0.022 (0.015) 0.057 (0.049) 0.039 (0.021)* 0.168 (0.063)*** 0.023 (0.017) Farm size

• 0.461

(0.171)***

• 0.535

(0.044)***

• 0.680

(0.214)***

• 0.520

(0.044)***

• 0.538

(0.076)***

• 0.586

(0.056)***

• 0.558

(0.136)***

• 0.528

(0.049)*** Share

• f

irrigated plots 2.588 (0.924)*** 3.497 (0.360)*** 4.538 (2.471)* 3.652 (0.346)*** 2.618 (1.365)* 3.665 (0.418)*** 1.949 (1.808) 2.935 (0.429)*** Family labor 0.905 (0.219)*** 1.027 (0.041)*** 1.052 (0.198)*** 1.029 (0.038)*** 0.649 (0.369)* 1.131 (0.048)*** 1.109 (0.203)*** 0.947 (0.057)*** Hired labor 0.452 (0.097)*** 0.239 (0.027)*** 0.319 (0.257) 0.249 (0.027)*** 0.318 (0.052)*** 0.265 (0.039)*** 0.387 (0.141)*** 0.285 (0.027)***

SLIDE 17

Results (3)... ..

 Endogenous r regime s swit itching r regressi sion m model l – Second s stage b by t type o

• f

f credit c constraint nts

Quantity constrained only Price constrained only Risk constrained only Transaction cost constrained

• nly

Constrained Unconstrained Constrained Unconstrained Constrained Unconstrained Constrained Unconstrained Use

• f

improved seeds 0.008 (0.336) 0.302 (0.094)***

• 1.002

(1.198) 0.331 (0.088)*** 0.254 (0.466) 0.435 (0.123)*** 1.907 (0.553)*** 0.286 (0.091)*** Use of fertilizer 0.352 (0.448) 0.779 (0.145)*** 0.829 (0.871) 0.751 (0.141)*** 0.811 (0.241)*** 0.875 (0.192)*** 0.829 (0.558) 0.437 (0.169)*** Use

• f

agro- chemicals 0.579 (0.336)* 0.212 (0.080)***

• 0.397

(0.514) 0.135 (0.078)*** 0.273 (0.142)*

• 0.005

(0.106) 0.766 (0.403)* 0.151 (0.083)* Average rainfall 2.101 (1.189)* 2.310 (0.367)*** 2.447 (0.689)*** 2.228 (0.097)*** 3.102 (0.481)*** 2.579 (0.276)*** 2.039 (0.412)*** 1.948 (0.147)*** Annual precipitation 0.933 (0.891) 0.065 (0.259) 1.822 (3.644)

• 0.037

(0.248) 2.311 (1.363)* 0.546 (0.365) 1.116 (1.166) 0.494 (0.288)* Mills ratio 1.864 (0.009)*** 0.228 (0.002)*** 2.269 (0.008)*** 0.094 (0.001)*** 0.900 (0.004)*** 0.746 (0.004)*** 1.735 (0.005)*** 0.232 (0.002)*** Constant

• 19.139

(10.369)*

• 17.819

(2.269)***

• 19.855

(29.080)

• 20.022

(2.178)***

• 6.105

(9.953)

• 24.536

(2.943)***

• 41.548

(9.526)***

• 15.165

(2.546)***

SLIDE 18

Results (4)

 Productivity gains fr from removing credit c constrain ints

Type of credit constraint A Percent in the sample B ( ) C ( ) D= B-C ( ) E=D/C Relative productivity gain Quantity-constrained 4.97 3,426.56 (6029.748) 2,983.058 (3,847.658) 443.502 14.9% Price-constrained 4.48 3,446.119 (10,136.96) 2,158.31 (5,278.067) 1,287.809 59.7% Risk-constrained 47.97 4031 (12,149.11) 2,882.7153 (1,529.69) 1,148.2847 39.83% Transaction costs-constrained 9.55 2,641.084 (10,603.38) 2,068.571 (4780.54) 572.513 27.68% All credit constrained households 66.64 5,023.002 (12,975.38) 3,133.034 (9,637.121) 1,889.968 60.03%

SLIDE 19

Conclusion and policy impli lications

 Main in fin indin ings

• Likelihood of accessing to and demanding credit was

significantly correlated with both supply- and demand-side factors

• Borrowing decision is influenced by presence of a

microfinance institution in the neighbourhood

• Determinants of credit constraints and their impact on farm

productivity are specific to the type of constraints farmers face

• The value of output per acre of constrained farmers could be

increased by more than 60% relative to the current level,

• nce all credit constraints have been removed

SLIDE 20

Conclusion and policy impli lications…

 Policy implications

• Tailored interventions on both supply and demand sides
• f credit markets
• Need to address insurance market failures in Ethiopia

given the prevalence of risk factors in explaining credit constraints

• Providing farmers with knowledge on financial services
• Increasing the number of credit providers, particularly

in rural areas, to reduce transaction costs

SLIDE 21

Thank y you for y your attention