Alternative Tracking Methods for Household Expenditure Ravi P. Rannan-Eliya Institute for Health Policy (IHP) Sri Lanka Consultation on National Health Accounts & External Resource Flows Ellison Institute for World Health Alexandria, VA July 21-22, 2005
21
Embed
Alternative Tracking Methods - IHP · Alternative Tracking Methods ... Ravi P. Rannan-Eliya Institute for Health Policy (IHP) ... Thailand, Malaysia Traditional providers, dentists,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Alternative Tracking Methodsfor Household Expenditure
Ravi P. Rannan-EliyaInstitute for Health Policy (IHP)
Sri Lanka
Consultation on National Health Accounts& External Resource Flows
Ellison Institute for World HealthAlexandria, VA
July 21-22, 2005
1
Problems of Household SurveyData
Example: Sri Lanka
2
Deaton (2003)
3
Recall loss for hospitalizations
4
Recent Estimation ExperienceProblems with household survey data
• Suffer from significant non-sampling errors• Eg: recall loss, telescoping, etc.
• Sampling errors if complete population is not covered• Existing surveys may lack sufficient detail to meet SHA
requirements• Surveys are costly, not timely, and are not available every year
Emerging approach• More intensive use of supply-side data and other projection
Private hospitalsNational surveys of hospital revenues
USA, Bangladesh, Sri Lanka, Hong Kong, MongoliaTriangulation from insurance data
Australia, Taiwan
10
Existing Methods (2)Private physicians
Tax data universally unreliable (including USA)Price * Quantity (PQ) approach
Use different surveys/sources to estimate P and QseparatelyHong Kong, Sri Lanka, Thailand, Malaysia
Traditional providers, dentists, paramedicalsSpecial surveys of providers, insurance data
Hong Kong, Korea, JapanTriangulation from household survey data
Calibrating using other reliable components
11
Combining methods in composite trendestimate
Focus on estimating elements as time series, notsingle cell estimates (temporal consistency)Requirement that estimates be consistent withdata sources at multiple timesEmphasis on accuracy of trends, not just levelsFocus on understanding trends, not single yearcross-section
Existing Methods (3)
12
Household Spending in Hong KongPublic sector user feesGovernment hospital data
Private hospitalsSpecial hospital survey,drug sales data
Duplicate measurement in theoryProblem: To estimate private clinic doctors revenues
14
Duplicate measurement in practiceProblem: To estimate private clinic doctors revenues
15
Duplicate measurement in practiceProblem: To estimate private clinic doctors revenues
16
Duplicate measurement in practiceProblem: To estimate private clinic doctors revenues
17
Duplicate measurement in practiceProblem: To estimate private clinic doctors revenues
18
Trend Estimation ExampleHong Kong DHA: Comparison of estimates of private
doctors' revenues
-
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
9,000
10,000
1989 1990 1991 1992 1993 1994 1995 1996 1997
GHS 1992 GHS 1996 IMS and tax data
HES 90/95 Proposed DHA estimate
Example: Hong Kong
19
IssuesCookbooks partially available (WHO PG, EurostatGuide)
Priority is to document existing sources outside OECDLack of OECD-type process outside OECD to improvemethods through cross-national discussion
Cookbook not sufficientExact methods will vary between country and over timeRequires capacity to evaluate, select and modify methodscontinuously
Scarcity and hidden nature of dataOften best overcome through multiple-year estimatesreconciling different data methodsDifficult to do well from outside country
20
Agenda for future
Regional and global stock-taking ofmethods
Currently not fundedMeeting-based mechanisms for expertsoutside OECD to share, review andimprove methods on routine basis