Transcript UNECE OECD EUROSTAT Group of Experts on National Accounts

UNECE OECD EUROSTAT Group of
Experts on National Accounts
Expanded Accounts for the
Household Sector
Discussant comments
Michael Connolly
May 2014
It’s about households
OECD
What’s going on within
the household sector?
Statistics Netherlands
Common Themes
• Papers highlight limitations of the existing statistics
when trying to understand h/holds
• Not without its challenges
– Adjustments for imputations by quintile
•
•
•
•
•
Underreporting by quintile
NL and OECD follow similar approach
Method A, B C (data, imputation, pro rata)
Calibration of overall micro totals to macro totals
NL present distribution across h/hold composition
rather than OECD presentation of quintiles
Overview OECD
• Omnibus paper
• Breakdown of Household Income, consumption,
saving and Wealth
• Link micro with macro – enhanced integration
• To allow distributional analysis across households
in the economy - understand impact of policy
– Quintiles
– Sources of income
– Household categories
Rationale - It’s all about inequality
• Stiglitz Sen Fitoussi
• Varying degrees of economic wellbeing across
households
• Uneven distribution of income
• Uneven ability to absorb economic shocks
Comparison of micro
data and National
Accounts totals
Distribution of
income consumption
and saving
OECD – Step 1 - Measurement and
comparison of Income Consumption
and Wealth
• Use of Adjusted Disposable Income - incl
transfers in kind
• Use of a Gap indicator as a measure of
mismatch
• Gap of 36% for ADI - after taking into account
quantifiable differences - gap reduced to 18%
• Significant divergence in either measure
• Imputations in SNA, under coverage in micro
surveys sub populations
OECD - Step 1 - Comparison of micro
data and National Accounts totals
• Micro- macro comparisons between 80% - 120%
OK
• Interest - Distributed income of Corporations and
income from self employed - poor
• Seem to be comparing different years for
different countries together
• Comparison between National micro source and
IDD may also relate to different years - eg saving
for 2006/7: NZ -3% and 2009/10: +16% AU
Step 2 - Distribution of Income,
Consumption and Savings
• 3 breakdowns - Income quintiles,
– main sources of income
– and household types
– 16 countries for 2008/2009/2010
• Inconsistencies in results for Income and consumption
– flatter profile for consumption
• Negative savings is the consequence
• Review of PIH and LCH : students and pensioners in 1st
quintile
• No conclusive results use of H of H/Hold data
• Non - observed income excluded from micro?
Statistical Issues/Dashboard
•
•
•
•
•
•
•
6 indicators
ADI – and GDP
COE & Mixed Inc of h/hold sector %GDP
Income redistribution
Net saving rate
H/Hold indebtedness
Unemployment
Netherlands Paper
• Rather than quintile analysis – Household category
– Age of head h/hold
– Household composition : couple with children, single
woman etc
– Standard of living
– Main source of income
• Disparity index – detail of analysis (h/hold
composition)
• Ratio Highest to lowest
• Overall directions of inequality – for Income and
consumption is down but up for wealth
Ageing Society
•
•
•
•
•
•
•
•
This is not an issue for Households but individuals
Generational perspective - level of individual
National Transfer Accounts
Wealth not considered
Doesn’t appear to make use of Pension Claims statistics
Concepts of income and consumption differ somewhat
HBS difficult to align to individual members of H/hold
How are transfers between h/holds recorded
Ageing society
• Link known average consumption expenditure
to h/holds classified by h/hold composition
• Allocate average amount across members of
h/hold using an equivalence scale depending on scale used different results
obtained
• Importance of flows within h/holds
• Linking between registers enabled h/hold –
individual , approach used in both projects
Questions for Authors
• Par 26 & 30 NL paper “disposable income and wealth is
greatest for couples with children”
• Although there appears to be some data linking in NL paper is there more scope?
• Statistical or economic analysis is this appropriate for NSIs ?
• Conclusions are seemingly tentative / experimental what is
required to improve quality of estimates?
• Would it be better to concentrate on households exclusively
rather than individuals / households as Institutional Units?
• Both approaches present the results differently – NL allows
some time series perspective but loses cross section analysis
by quintile – OECD vice-versa are there other presentational
proposals
EEA Household Sector
Accounts
ECB
NPISH in the Canadian
System of National Accounts
Statistics Canada
Discussants comments - EEA
• Steep learning curve for non-expert users !!
• Very useful overview of Single Currency Area
• Highlights incipient household problems property/lending
• Overall integrated framework - do we miss
the non-financial balance sheets and
valuation?
Figure 3 Savings, Investment and Net Lending/Borrowing of Households
€billions
50
40
30
20
10
0
-10
-20
-30
-40
2002
2003
GFCF
2004
2005
Net Capital Transfers
2007
2006
Shares
2008
Deposits
2009
Loans
2010
2011
Insurance&Pensions
2012
Gross Savings
Canadian NPISH Accounts - comments
• Excellent practical description of how to compile
NPISH sequence of accounts
• Emphasis on starting point - establish the
universe - vital
• Data sources for Register building informative
• The key elements of NPISH expenditure and
income are highlighted
• Good detail on data sources - similar to other
countries
Questions
• Is the financial corporations element of what was
“persons and unincorporated businesses sector”
excluded from the new NPISH - presumably “yes”
• There is an increase in Financial assets as a proportion
of total assets and a decrease in non-financial
assets???
• Is the establishment of foundations for receipt of
donations /govt. transfers regulated
• In relation to the increased contribution of Govt.
transfers to NPISH income - are these increased
transfers concentrated in areas like health/eduction?