Empirical Project 9 Solutions

These are not model answers. They are provided to help students, including those doing the project outside a formal class, to check their progress while working through the questions using the Excel or R walk-throughs. There are also brief notes for the more interpretive questions. Students taking courses using Doing Economics should follow the guidance of their instructors.

Part 9.1 Households that did not get a loan

Region Proportion in the region living in small towns Proportion of regional households living in large towns Proportion of regional households living in rural areas Proportion of all households living in the region
Addis Ababa 0 1.00 0 0.06
Afar 0.15 0.10 0.76 0.03
Amhara 0.11 0.22 0.67 0.20
Benshagul Gumuz 0.10 0.00 0.90 0.02
Diredwa 0.00 0.47 0.53 0.04
Gambelia 0.08 0.12 0.80 0.02
Harari 0.00 0.27 0.73 0.03
Oromia 0.11 0.28 0.61 0.20
SNNP 0.09 0.18 0.73 0.23
Somalie 0.09 0.16 0.76 0.06
Tigray 0.07 0.37 0.56 0.12
Grand total 0.09 0.28 0.63 1.00

Proportion of sample living in towns vs rural areas, by ‘region’. (Note that numbers may not add up to 1 due to rounding.)

Solution figure 9.1 Proportion of sample living in towns vs rural areas, by ‘region’. (Note that numbers may not add up to 1 due to rounding.)

  Mean Standard Deviation Min Max
Household size 4.58 2.40 1.00 16.00
Gender 0.30 0.46 0.00 1.00
Age 44.18 15.61 3.00 99.00
Young children 1.89 1.71 0.00 10.00
Working-age adults 2.58 1.52 0.00 10.00
Max education 7.53 7.28 0.00 30.00
Number of assets 14.90 17.23 0.00 203.00

Summary table for household demographics (all households).

Solution figure 9.2 Summary table for household demographics (all households).

Not rejected Rejected Blank or NA Sum
Applied for loan 1,363 201 1 1,565
Did not apply for loan 3,632 24 2 3,658
Blank or NA 37 2 0 39
Grand total 5,032 227 3 5,262

Loan applications and approvals.

Solution figure 9.3 Loan applications and approvals.

Reason Proportion
Do Not Like To Be In Debt 0.19
Have Adequate Farm 0.19
Fear Not Be Able To Pay 0.17
Believe Would Be Refused 0.12
No Farm or Business 0.10
Do Not Know Any Lender 0.07
Too Expensive 0.05
Inadequate Collateral 0.05
Other (Specify) 0.04
Too Much Trouble 0.03

Most important reason for not applying for a loan.

Solution figure 9.4 Most important reason for not applying for a loan.

Reason Proportion
Fear Not Be Able To Pay 0.28
Do Not Like To Be In Debt 0.24
Inadequate Collateral 0.09
Believe Would Be Refused 0.09
Too Expensive 0.07
Do Not Know Any Lender 0.06
Too Much Trouble 0.06
Have Adequate Farm 0.05
No Farm or Business 0.04
Other (Specify) 0.02

Second most important reason for not applying for a loan.

Solution figure 9.5 Second most important reason for not applying for a loan.

  1. Solution figure 9.6 compares the loan purposes of ‘denied’ and ‘successful’ borrowers. Compared with all households, a greater proportion of households that got loans indicated the purpose for the loan as consumption, and a smaller proportion of households that got loans indicated the purpose as for investment.

    Note: The entry ‘10’ is likely due to incorrect recording of responses, since ‘10’ is not a valid category in the original survey.

Purpose Denied Successful
10 0.02 0.00
Business Start-up Capital 0.26 0.15
Expanding Business 0.14 0.08
Other (Specify) 0.26 0.03
Purchase Agricultural Inputs for Food Crop 0.21 0.30
Purchase House/Lease Land 0.03 0.02
Purchase Inputs for other Crops 0.06 0.01
Purchase Non-farm Inputs 0.03 0.12
For consumption and personal expenses 0.00 0.20

Loan purpose for households denied a loan (credit-excluded) and successful households.

Solution figure 9.6 Loan purpose for households denied a loan (credit-excluded) and successful households.

Note

Although most categories are the same, the two tables are not exactly comparable because ‘Consumption and personal expenses’ was a separate category from ‘Other’ for successful borrowers but was considered as one of the responses in ‘Other (specify)’ for credit-excluded households. Looking at the variable loan_purpose_other in the ‘All households’ tab, we can see that reasons related to consumption and personal expenses (such as medical, school, transport expenses) apply to around 20 or so households.

Household characteristic Successful borrowers Denied borrowers
Number of observations 1,363 201
Age of household head 43.40 41.20
Highest education in household 7.26 8.00
Number of assets 15.90 14.50
Household size 4.87 4.82
Number of young children 2.09 2.22
Number of working-age adults 2.75 2.76

Comparison of household characteristics.

Solution figure 9.7 Comparison of household characteristics.

  Rural   Small town (urban)   Large town (urban)  
  Successful Denied Successful Denied Successful Denied
Number of observations 903 128 128 17 332 56
Age of household head 45.10 43.40 42.10 37.50 39.20 37.30
Highest education in household 5.00 4.55 9.59 13.12 12.51 14.3
Number of assets 13.60 12.10 15.70 16.70 22.10 19.20
Household size 5.35 5.47 4.12 4.06 3.84 3.57
Number of young children 2.49 2.84 1.56 1.41 1.20 1.05
Number of working-age adults 2.91 2.88 2.54 2.33 2.42 2.36

Comparison of household characteristics, conditioning on the ‘rural’ variable.

Solution figure 9.8 Comparison of household characteristics, conditioning on the ‘rural’ variable.

Successful borrowers Denied borrowers Difference in means
Household characteristic Mean SD N Mean SD N Difference in Means CI Lower CI Upper
Age of household head 43.40 14.30 1,361 41.20 12.90 201 2.15 0.21 4.09
Highest education in household 7.26 6.74 1,361 8.00 7.90 201 −0.73 −1.89 0.42
Number of assets 15.90 19.00 1,361 14.50 16.80 201 1.42 −1.13 3.97
Household size 4.87 2.35 1,361 4.82 2.35 201 0.05 −0.30 0.40
Number of young children 2.09 1.68 1,361 2.22 1.80 201 −0.14 −0.40 0.13
Number of working-age adults 2.75 1.49 1,361 2.76 1.42 201 −0.003 −0.22 0.21

Calculating 95% confidence interval for difference in means.

Solution figure 9.9 Calculating 95% confidence interval for difference in means.

Differences in household characteristics, with 95% confidence intervals.

Solution figure 9.10 Differences in household characteristics, with 95% confidence intervals.

Household characteristic Successful borrowers Denied borrowers Discouraged borrowers Constrained borrowers
Number of observations 1,363 201 588 3,012
Age 43.37 41.21 43.28 44.84
Highest education in household 7.26 8.00 6.50 7.14
Number of assets 15.88 14.46 10.16 13.54
Household size 4.87 4.82 4.65 4.44
Number of young children 2.09 2.22 2.03 1.81
Number of working-age adults 2.75 2.76 2.49 2.49

Comparison of household characteristics.

Solution figure 9.11 Comparison of household characteristics.

  1. One example of selection bias is the study of the determinants of prosperity. To do this, we can compare developing with developed countries. However, only developing countries with adequate resources and willingness collect and submit their data. The set of developing countries for which we have data is thus not representative of the population of developing countries.

Part 9.2 Households that got a loan

  Mean SD Min Max
Loan amount (principal) 26,896 783,587 1 30,000,000
Total amount 29,223 827,144 20 31,260,000

Summary measures of loan amount (principal) and total amount.

Solution figure 9.12 Summary measures of loan amount (principal) and total amount.

Loan amounts and interest rates.

Solution figure 9.13 Loan amounts and interest rates.

Loan amount N Mean SD Min Max 1st quartile 2nd quartile 3rd quartile
Long-term 211 172,718 2,072,736 20 30,000,000 1,000 3,700 8,000
Short-term 717 3,017 8,398 40 150,000 480 1,500 3,500
Interest rate N Mean SD Min Max 1st quartile 2nd quartile 3rd quartile
Long-term 211 0.19 0.27 0.00 2.24 0.00 0.14 0.25
Short-term 717 0.11 0.17 0.00 1.12 0.00 0.05 0.17

Comparison of distribution of long-term and short-term loans.

Solution figure 9.14 Comparison of distribution of long-term and short-term loans.

Short-term loans are, on average, for smaller amounts. A greater proportion of short-term loans offer very low interest rates. Smaller proportions of short-term loans offer high interest rates. The spread of interest rates and loan amounts is much larger for long-term loans than for short-term loans.

Household characteristics Interest rate
Household size 0.11
Gender –0.02
Age 0.03
Young children 0.10
Working-age adults 0.05
Max education −0.08
Number of assets –0.05

Correlations between interest rate and household characteristics.

Solution figure 9.15 Correlations between interest rate and household characteristics.

The correlations are all fairly weak. We would expect that lenders would charge lower interest rates to households with more working-age adults and more members because these households are better able to repay, and the risks associated with lending to them are therefore lower. This is not what we observe in the data. The rest of the correlations are as expected.

Proportion of 'borrowed_from'
Source of finance Large town (urban) Rural Small town (urban) Total
Bank (commercial) 0.02 0.00 0.00 0.01
Employer 0.04 0.00 0.01 0.01
Grocery/Local Merchant 0.08 0.05 0.10 0.06
Microfinance Institution 0.19 0.28 0.27 0.26
Money Lender (Katapila) 0.00 0.05 0.02 0.04
Neighbour 0.11 0.12 0.07 0.11
NGO 0.01 0.05 0.05 0.04
Other (specify) 0.06 0.12 0.04 0.10
Relative 0.49 0.32 0.43 0.37
Religious Institution 0.00 0.02 0.00 0.02

Source of loan, by variable ‘rural’.

Solution figure 9.16 Source of loan, by variable ‘rural’.

Proportion of borrowed_from_other
Source of finance Large town (urban) Rural Small town (urban) Total
Bank 0.00 0.01 0.00 0.01
Cooperatives 0.42 0.43 0.40 0.43
Equib 0.05 0.01 0.00 0.01
From government 0.05 0.30 0.00 0.26
From individuals 0.05 0.01 0.00 0.01
From private business 0.00 0.02 0.00 0.02
From relatives 0.05 0.00 0.20 0.01
From women association 0.00 0.01 0.00 0.01
From Youth Association 0.00 0.01 0.00 0.01
HAB project 0.00 0.02 0.00 0.01
Iddir 0.16 0.15 0.00 0.14
Micro and small enterprise 0.11 0.00 0.00 0.01
Micro finance 0.00 0.04 0.40 0.05
Mobile 0.05 0.00 0.00 0.01
NGO 0.05 0.00 0.00 0.01

Source of loan, by variable ‘rural’ (‘Other’ category only).

Solution figure 9.17 Source of loan, by variable ‘rural’ (‘Other’ category only).

Compared to rural households, urban households are more likely to borrow from banks and employers. Rural households are more likely to borrow from microfinance institutions and governments.

Source of finance Large town (urban) Rural Small town (urban)
Bank (commercial) 1,814 619  
Employer 602 290  
Grocery/Local Merchant 166 259 176
Microfinance Institution 712 411 510
Money Lender (Katapila) 365 332 365
Neighbour 125 187 296
NGO 373 395 236
Other (specify) 274 372 806
Relative 237 217 393
Religious Institution 1,461 343  
(blank) 1,096 289 151

Duration of loan (rounded to nearest day).

Solution figure 9.18 Duration of loan (rounded to nearest day).

Source of finance Large town (urban) Rural Small town (urban)
Bank (commercial) 5,755,833 3,575  
Employer 5,915 1,350 1,000
Grocery/Local Merchant 1,370 1,429 1,599
Microfinance Institution 14,525 3,852 7,543
Money Lender (Katapila) 350 1,360 5,150
Neighbour 602 837 686
NGO 5,532 1,4475 2,300
Other (specify) 3,912 1,842 3,829
Relative 7,872 1,576 6,802
Religious Institution 50,000 910  

Loan amount (rounded to nearest whole number).

Solution figure 9.19 Loan amount (rounded to nearest whole number).

Source of finance Large town (urban) Rural Small town (urban)
Bank (commercial) 0.14 0.26  
Employer 0.04 0.06 0.00
Grocery/Local Merchant 0.00 0.09 0.00
Microfinance Institution 0.16 0.18 0.15
Money Lender (Katapila) 1.00 0.43 0.09
Neighbour 0.00 0.12 0.00
NGO 0.06 0.12 0.05
Other (specify) 0.13 0.16 0.22
Relative 0.01 0.08 0.00
Religious Institution 0.00 0.20  

Interest rate (rounded to two decimal places).

Solution figure 9.20 Interest rate (rounded to two decimal places).

Source of finance Large town (urban) Rural Small town (urban)
Bank (commercial) 0.00 0.50  
Employer 0.31 0.00 0.00
Grocery/Local Merchant 0.39 0.19 0.50
Microfinance Institution 0.41 0.13 0.31
Money Lender (Katapila) 0.00 0.27 0.50
Neighbour 0.35 0.25 0.43
NGO 0.25 0.31 0.20
Other (specify) 0.39 0.19 0.25
Relative 0.40 0.21 0.29
Religious Institution 1.00 0.24  
(blank) 1.00 0.43 1.00

Proportion of households with a female head, according to source of finance.

Solution figure 9.21 Proportion of households with a female head, according to source of finance.

Households with female heads are relatively more likely to borrow from grocery/local merchants, neighbours, NGOs, and relatives.

  1. One hypothesis is that lack of knowledge of loan availability affects access by households to lending. The government could fund education programs aimed at improving the poor’s understanding of financial services. The government could identify two poor regions with similar characteristics that are distant from each other. One region would serve as the control group while the other as the treatment group. The government could then randomly select and educate individuals from the treatment region. The causal effects of the policy could be assessed by comparing outcomes in the two regions after the treatment. During the study period, the government should avoid implementing other policies in the regions, especially those that can affect the regions differently. The two regions should be chosen such that they would evolve in similar ways without the policy, and that treatments on one region cannot indirectly affect outcomes in the other region.