ISVEE16 – Halifax, Canada
16th International Symposium of Veterinary
Epidemiology and Economics (ISVEE 16)
Connecting Animals, People, and their shared
environments
Estimating the effects of a change in dairy policy on infants’
milk intake in Kenya and Tanzania
Emmanuel Muunda1, Nadhem Mtimet2, Francis Wanyoike1, Paula Dominguez-Salas3, Silvia Alonso1
1International Livestock Research Institute
2International Fund for Agricultural Development
3Natural Resources Institute, University of Greenwich
22nd International Symposium of Veterinary Epidemiology and Economics
Halifax, Canada, 9 August 2022

ISVEE16 – Halifax, Canada
Kenya & Tanzania Dairy Sectors
Dairy Consumption
Formal Dairy
 Controls 70 - 80% of milk
market
 Largely unregulated
 Markets unprocessed
milk
 Cheaper by 20-50%
 Sold in variable quantities
depending on
affordability
 Preferred for taste, high
butterfat content, widely
accessible - Kenya
 Food safety concerns:
handling, adulteration,
lack of traceability, etc.
 Controls about 20 - 30% of
milk market
 Regulated
 Markets processed milk
 More expensive
 Better food safety & quality
status, traceability
measures
Informal Dairy
Kenya
 Per Capita annually –
110liters
 Child intake averages 300ml
(1 cup) per day
 EBF at 61% rest largely
given milk
Tanzania
 Per Capita annually – 47
liters
 Child intake no info found
 EBF rate is 40.8%

ISVEE16 – Halifax, Canada

ISVEE16 – Halifax, Canada
Policy Interventions
• Kenya: The Dairy Industry (Dairy Produce Safety) Regulations, 2021 : Part
IX – Milk bar
 Only selling pasteurized dairy produce; - restricts sale of raw milk
 Dispense dairy produce hygienically from a labelled, tamper proof, food
grade and easy to clean equipment
 Operator: persons who are capable of conducting the basic quality
assessments tests; keep records of the quality tests.
• Tanzania: The Dairy Industry Act, (Cap. 262), 2020:
 “A person shall not offer for sale, sell or supply unpasteurized
milk to any person unless such milk sold or supplied to milk
collection centres or milk processing factories Only selling
pasteurized dairy produce”

ISVEE16 – Halifax, Canada
The choice experiment
• The choice experiment posed 9 hypothetical scenarios of 4 choices each
Attributes/Alternatives
A1 Decrease raw milk quantities for all family members without replacing it by any other food product
A2
Decrease raw milk quantities for all family members, and replace it with another food product only
for children <4 years
A3
Decrease raw milk quantities for all family members, and replace it with another food product for
all family members except for children <4 years
A4
Decrease raw milk quantities for all family members, and replacing it with another food product for
all family members
A5
Keep raw milk quantities the same for children < 4years and decrease it for the rest of family
members
A6
Decrease the quantities of raw milk I give to the children <4 years, without replacing it by other
food products. Will keep the same quantities of raw milk for adults
A7
Decrease the amount of raw milk I give to the children <4 years, while replacing it by other
products. Will keep the same amount of raw milk for adults
A8 Keep buying the same quantities of raw milk by increasing milk budget
A9 Stop buying raw milk

ISVEE16 – Halifax, Canada
S03q01. If raw milk price increases by 40% compared to high season prices, which corresponds to new raw milk
price around KES 100/litre, from the 4 alternatives/actions below please indicate which is the most likely
alternative/action you will choose/do and the least likely alternative/action you will not choose/do? (Tick only one
case as most important and one case as least important)

ISVEE16 – Halifax, Canada
Analytical Approach
 Best-Worst Scores – Relative importance of choices
 Mixed Logit Model: To confirm the ranking above and identify
heterogeneity in choices
 Latent class model: Latent class analysis groups cases or
scenarios into classes or categories:
(Latent class analysis is a statistical method for identifying unobserved class
membership among subjects using categorical and/or continuous observed
variables)
Heterogenic groups & - Homogeneity within group
 Use HH socioeconomic factors to characterize groups

ISVEE16 – Halifax, Canada
Relative Importance of the choices
Kenya – Best Tanzania – Best
 3 of 4 reduced intake
 3 substitute for children
 2 of 4 reduced intake
 2 substitute for kids
33%
30%
35%
23%
0.8% 0.2%
0.8%
1%
& Worst & Worst

ISVEE16 – Halifax, Canada
Latent classes – In both, 3 classes
 Class 1 - Kenya – A4, A2 & A7 have the highest coefficients; the
quantities of milk allocated to children decreases and is replaced
with another food item. (TZ – same as Kenya)
 Class 2 – A8, A2 & A5 have the highest coefficients; the most
important alternative is A5 which represents keeping the raw milk
quantities allocated to children and decrease for the rest of
family.
(TZ – A5, A2, A8)
 Class 3 – A2, A4 & A5 - lower estimation magnitudes. A3 & A6
are not statistically different from the reference level (choice-A9)
in both.(TZ – A1, A6, A3)

ISVEE16 – Halifax, Canada
Composition of groups - Kenya
Parameter Class 1(64%) Class 2(22%) Class 3(14%)
Household Income**(USD)
Below 100 16.92 19.05 28.57
101-200 39.23 26.19 42.86
201-300 43.85 54.76 28.57
Total 100 100 100
Gender of HH Head
Male 76.92 85.71 89.29
Female 23.08 14.29 10.71
Total 100 100 100
Age of HH head***
18 - 29yrs 37.69 38.1 25
30 - 39yrs 43.85 40.48 53.58
40 - 49yrs 13.08 16.67 10.71
50yrs and Above 5.38 4.75 10.71
Total 100 100 100
Education level of HH Head*
Primary / Vocational school 29.46 42.50 28
Secondary school (form 1-4) 44.96 47.5 60
Technical/University 25.58 10 12
Total 100 100 100
Mean Raw Milk Expenditure (KES)* 313.84a,b 235.73a 205.18b
Mean Quantity of raw milk purchased
(liter)
4.00a 3.46 2.70a
Number of children (6 – 48months old)** 1.19 1.12 1.04
Household size (Mean) 4.36 4.33 4.17

ISVEE16 – Halifax, Canada
Composition of groups - Tanzania
Household variable Level
Class 1 (30.2%) Class 2 (27.8%) Class 3 (42%)
Household Income (%)* USD52 and Below 46.67 32.73 34.48
Between USD53 and USD86 31.67 38.18 37.93
Between USD87 and USD172 21.67 29.09 27.59
Gender of HH head (%)
Male 80.00 78.18 88.37
Female 20.00 21.82 11.63
Activity of HH head (%)**
Self-employed (own/family business) 83.33 80.00 77.91
Employed/labourer 8.33 12.73 22.09
Unemployed 8.33 7.27 0.00
Age of the HH head (%)*
18 - 29 years 16.67 16.36 22.09
30 - 39 years 23.33 38.18 32.56
40 - 49 years 21.67 14.55 25.58
50 Years and Above 38.33 30.91 19.77
Highest education Level of HH head
Adult literacy education 0.00 0.00 1.35
Primary school (class 1-8) 88.24 75.56 79.73
Vocational school (no secondary
education)
0.00 4.44 0.00
Secondary school & above 11.76 20.00 18.92
Mean monthly food expenditure (TSh) 138915a 133181.8a,b 111034.5b
Mean monthly non-food expenditure (TSh) 72694.17a 41138.18b 32550.57b
Number of children below 48months 1.2a 1.2a 1.2a
Mean Household size 6.1a 5.8a,b 5.2b

ISVEE16 – Halifax, Canada
Policy implication
 Given the evidence that overall demand for milk is decrease with
increased price, dairy policies should consider milk affordability in
order to safeguard nutrition security of children. This may involve
interventions that increase production and strengthening the supply
chains
 There is a need to strengthen resilience to milk price variations in
poor households. Considering that a bigger proportion of the
respondents preferred replacing milk with other food items, often
fruits, there is a need to identify and create public awareness on
food substitutes that offer similar or better nutritional value as milk
at similar or lower price and preparation costs. But do such food
substitute with these specifications exist?

ISVEE16 – Halifax, Canada
Policy implication (Cont..)
 Low-income consumers represent the largest segment of
the Kenya population and thus are the biggest milk
consumers (in total vol) of milk. The study showed that
these consumers are price sensitive and that the
increase in milk prices will reduce their milk purchase and
the quantities allocated to their infants (less than 4 years
old). This will have negative impacts on low-income
household infants’ nutrition in Kenya.

ISVEE16 – Halifax, Canada
Progressive Policy: 3P Summary
Profits(Livelihoods, incentives)
People(Consumers, nutrition)
Processing (Food safety, regulations)
Fair & Competitive dairy markets
Selling safer milk & milk products
Contributes to nutrition needs of the poor, especially children

ISVEE16 – Halifax, Canada
Read more…
https://doi.org/10.1016/j.foodpol.2021.102043

ISVEE16 – Halifax, Canada
Acknowledgements: MoreMilk project
• Donors • Project partners

THANK YOU