1. HOUSEHOLD INCOME AND EXPENDITURE SURVEYS IN THE PACIFIC REGION
Pacific Statistics Methods Board,
Nadi, Fiji,
30 October, 2018
Michael Sharp, Bertrand Buffiere, and Olivier Menaoeur,
Statistics for Development Division

2. Purpose Invite the Pacific Statistics Methods Board (PSMB) to:
Note the history of HIES in the Pacific region;
Note the current regionally standardised HIES methodology;
Note the main challenges in conducting HIES under the Pacific “regionally standardised” methodology;
Note the HIES methodological experiment that is currently being conducted in RMI; and
Endorse the recommendation for PMSB to review the results of the RMI HIES experiment at PMSB’s third meeting to be held in 2019 in order to make an informed recommendation on the optimal methodology to conduct HIES.

3. Introduction to HIES

4. HIES objective and use
Multi-purpose and multi-user survey with the objective to:
Update the household component of GDP
Rebase the CPI
Conduct poverty analysis
Conduct general demographic and socioeconomic analysis
More recently, conduct food security and nutrition oriented analysis
Directly populate 17 of the 132 Pacific subset of SDG indicators (13%)
Contributes data to an additional 6 indicators (17% of the total subset)

5. History of HIES in the Pacific region

6. History of HIES in the Pacific
70 HIES conducted since 1943 (Fig1) with 60% since MDG era
Excluding Pitcairn, all PICTs conducted at least 1 HIES (Fig2)
3 PICTs with only 1 HIES
3 PICTs with most recent HIES pre-2010
All scheduled HIES in 2019
Desirable to have more, but costly (NSO, HH, TA provider, donor)
Using data from 11 recent surveys, average cost is USD 340 per enumerated household, or USD 10 per capita. Economies of scale is achieved in larger PICTs where the costs is spread over a larger population and sampling fractions are relatively low. The call for more frequent production of highly disaggregated data presents an immense challenge for Small Island States.

7. The regionally standardised HIES methodology

8. Pacific standardised HIES
Implemented in the Pacific since 2012
Developed under the guidance of a technical working group
Standardised:
Survey instruments
Classifications
Field operations schedule
Data capture and processing system
Reporting and data structure

9. HIES questionnaire
Module 1 - Demographic information: demographic profile; activities last week; education; health; communication; and previous household members.
Module 2 - Household expenditure: dwelling characteristics; dwelling tenure; utilities and communication; land and dwelling; goods and assets; vehicles; private travel; services; cash contribution to special occasions; provision of financial support; insurance and taxes; and loans.
Module 3 - Individual expenditure: education; health; clothing; communication; luxury items; and alcohol, betel nut, kava and tobacco; (optional: deprivation).

10. HIES questionnaire
Module 4 - Household and individual income: wages and salaries; agriculture and forestry; fishing, gathering and hunting, livestock and aquaculture; handicrafts and home processed foods; non-subsistence business; property, transfer income; and remittances and cash gifts.
Two-week diary: cash expenditure on food and non-food items; cash expenditure on services; goods and services received for free; home produced items today (home consumption, to give away and to sell).
Main methodological issues surround the use of the diary.

11. Field schedule Rolling 12-months sample (seasonality)
Field teams: Supervisor (supervision and CAFE data entry) and two enumerators
Two-week diary requires field teams to be in an EA for three-weeks
16 HIES rounds – main constraint
Seven to eight visits per household per round
Six to seven households per enumerator per round (~100 per enumerator in total)

12. Challenges with the regionally standardised HIES methodology

13. Methodological challenges
Despite the regionally standardized HIES meeting the multi-user needs, there are a number of inherent methodological issues, including:
High cost
Apparent respondent and/or enumerator fatigue
Diary versus recall
Food away from home
Food acquisition as a proxy for consumption
Non-standard units
Partakers
Imputed rents and consumption of fixed assets

14. High cost

15. Fatigue Literacy Time poor
Seven to eight enumerator visits recommended
Diary “drop off” in transactions over the two- week diary period (ref: Fig)
Well managed diary considered to be “gold standard”

16. Diary versus recall COICOP Division 2 (alcohol, tobacco and narcotics)
May be underreporting in the diary as COICOP Division 2 is already captures in the modules, but clear differences in expenditure when comparing diary and recall data sources

17. Food away from home Food acquired and consumed away from the dwelling
Anecdotally a significant calorie source for Pacific Islanders (feasts, takeaway)
No dedicated section in HIES questionnaire (Tonga 2009 v Tonga 2015)
Ref fig: Proportion of total food expenditure from COICOP Group 11.1
Solomon Islands “pocket money”
1,580 households (around 35 percent of sampled households) in the 2012 Solomon Islands HIES reported “pocket money” in the diary. This amounted to 5,860 transactions averaging SBD 41 per transaction and 6.7 percent of total cash expenditure reported in the diary. Whilst it’s possible that pocket money was used for transport and other non-food expenditure, it’s likely that a large proportion of pocket money was spent on food away from home.

18. Non-standard units of measurement
Common methodological issue across all HIES (or its equivalent)
Ref fig: Proportion of food items reported in NSU
Statistical methods available, but need critical mass
Market and producer surveys over time and geographical domain

19. Other methodological challenges
Food acquisition as a proxy for estimating consumption
Very high and very low food acquisition
Home production for consumption today or total home production
Opening and closing stocks (Consumption = opening stock + food acquisition – closing stocks)
Partakers
Food sharing is a major component of the Pacific culture
Partakers not well captured to account for sharing of food in consumption estimates
Imputed rents
No baseline rental market
House value and age not collected
Consumption of fixed assets
Age and value of durable goods not collected

20. The RMI HIES experiment

21. Acknowledgements

22. RMI HIES experiment
Response to the challenges and commitment to produce high quality data
Three main components:
Testing new technology (CAPI) and backup protocols in small island states
Inclusion of complementary survey modules
New methods for the collection of consumption data

23. New technology
CAPI (Survey Solutions software) of complex HIES questionnaire
Electrification
Internet connectivity (data backup)

24. Complementary survey modules
Dynamic data and avoid need for stand-alone surveys
Food away from home: recall and diary-based modules are being tested.
Partakers: recall and diary-based modules are being tested.
Labour module: newly designed module in accordance with the resolution concerning the International Classification of Status in Employment (ICSE).
Food Insecurity Experience Scale (FIES): to measure experiences of moderate-to-severe food insecurity and to report against SDG indicator
Shocks: household exposure to shocks through adverse environmental, economic and social change.
Disability: addition of the Washington Group six questions to identify vulnerable populations.
Improved modules: enhanced health and fisheries modules and improved means to capture information to better estimate household consumption of fixed assets.

25. Improved methods for the collection of consumption data
Diary v recall
PAPI v CAPI
Food away from home
Partakers
Does diary or recall produce more comprehensive consumption data and which implementation protocol will result in the efficient production of high quality consumption data?

26. Five experimental arms
Recall CAPI: household members are asked to recall if they acquired and consumed food and non-food items in the last seven days; they are asked the number of partakers in each meal in the last seven days; and they are asked if the household members acquired food away from home in the last seven days. Responses are entered in the tablets (CAPI).
Bounded recall CAPI: same as Recall CAPI, however the enumerator will visit the household seven days before conducting the recall interview to ask the household to try to remember what is acquired/consumed over the next seven day period leading up to the interview.
Diary CAPI – high monitored: household keeps a 14-day diary of all food and non-food items that the household acquires each day (household); they are also asked to record if they acquired and consumed food away from home (individual); the enumerator will visit the household every second day to enter the diary data into the tablet and to ensure that the household is completing the diary comprehensively.
Diary PAPI – high monitored: this is the same as Diary CAPI – high monitored, except the diary data is not entered into the tablet during the visit to the household; it is entered into a data entry system after each round; the enumerator will still visit the household every second day to ensure that the diary is being completed.
Diary PAPI – low monitored: this is the same as Diary PAPI – high monitored, however the enumerator will only visit the household to drop the diary off on day one, then pick up the first week diary on day seven and drop the second week diary, then pick up the second week diary on day 14. This is a poorly managed diary without regular monitoring of the household to ensure the diary is being completed.

27. RMI HIES experiment – next steps
Field work will finish in November and the results will be analysed and reported on up until April It is at this time when conclusions will be made as to the optimal methodology for conducting HIES in the Pacific region. In drawing of the conclusion, all data applications (e.g., for national accounts, CPI, poverty, nutrition and food security) will be taken into consideration to ensure that the new HIES methodology continues to meet its multi-user needs.
The results, conclusion and recommendation will be presented to PMSB in 2019 for endorsement.

28. Recommendation Pacific Statistics Methods Board is invited to:
Note the history of household income and expenditure surveys (HIES) in the Pacific region;
Note the current regionally standardised HIES methodology;
Note the key challenges in conducting HIES under the “regionally standardised” methodology in the Pacific region;
Note the HIES methodological experiment that is currently being conducted in the Republic of the Marshall Islands (“the RMI HIES experiment”); and
Endorse the recommendation for PMSB to review the results of the RMI HIES experiment at PMSB’s third meeting to be held in 2019 in order to make an informed recommendation to the Pacific Island countries and territories on the optimal methodology to conduct HIES.

29. Thank you