Demography and
economics: Matt Berkley, correspondence
and other documents 2000-3
With quotations from others
|
Matt Berkley |
Others |
|
policy-makers
need help to understand how economic analysis can reflect individual outcomes
among the most vulnerable. [1] ....is there any reason for the total mix
of a government's policies not to be evaluated on the number of deaths among
the most-vulnerable? [2] |
increasingly, policy makers have recognized that one part of
what it means to be poor resides in a sense of vulnerability to devastating
loss [3] |
|
over 100 million early
deaths [are surely projected by 2015]...If these people die, the income
figures look a lot better....One catalyst for this approach could be "we
need a rethink due to HIV and the digital divide”. [4] This
seems to me the worst flaw in simple economic analysis...[5] |
Many millions of people
around the world are going to die from AIDS ....(...If per capita income were
to rise as people died, would that make the impact of AIDS somehow less
severe?) My own view is that the Bank should be backing away from its current
too concentrated focus on the income headcount numbers....[6] |
|
[I would like to] get data which reflects the life courses of
individuals...[7] |
I like your life-cycle idea. [8] one fruitful avenue....a greater focus on life-cycle
income. [9] |
|
Many
economists think this is a minor technical issue. It is not: it is
a fundamental question of what is acceptable social science. [10] |
fixing
this problem is not simply a technical matter. [11] |
|
We need mortality data or cohort data
to ensure that incomes are being raised
[12] |
we
need to have data on birth dates (ti), length of life (li) and income
profiles [13] |
|
a family where there are more surviving children is
counted as “poorer”, but one
where....more died of malnutrition, looks “better off” [14]
Child deaths in a family result in a higher per capita income [15] |
a household whose young children die may
be reported as escaping poverty when a similar household that manages to
raise all its children ....is seen as ‘failing’ to escape from poverty as its
income/ consumption levels per capita are lower [16] |
|
Other things being
equal, the Bank must have overestimated the reduction in the proportion of
people living on the original level of per-day consumption considering
their age and size. A dollar a day,
even if we knew its value in food terms, would not be an appropriate measure
where the proportion of children is changing, as now. [17]
The World
Bank’s “dollar” line....mistakenly counts adults’ needs as the same as
children’s. [and] will underestimate poverty as birth rates fall...[18] other things being equal.....If the birth rate falls, and babies
earn less than adults, average income will go up even if average incomes for
all age groups stay the same. [19] |
[the per capita] approach
is taken, for example, in calculating the widely-used $1/day and $2/day per
capita poverty lines....a population experiencing a rapidly declining
fertility rate....will experience an exaggerated decline in the short-term
reduction in poverty when poverty is measured on a per capita basis [20] [No comment] [21] [No reply] Thank you. [22] |
|
Life
expectancy fell in Uganda in the 90s while the economic statistics did
better. Just looking at the economic
statistics would not be to measure aggregate welfare gains and losses. [23] |
when
life expectancy changes systematically and out of harmony with consumption
patterns, as it has in ...Africa in the 1990's, ignoring it in welfare
calculations will result in misleading inferences. [24] |
|
Sen also talks about capabilities....I
would like to see more of an emphasis on increasing capabilities by removing
constraints (the main one being risk of dying) since these are what matter
most to people. [25]
....any Benthamite definition of utility
includes the duration of pleasure, or happiness, or utility, or well-being,
or capabilities. [26] |
acknowledging
Sen’s arguments for extending well being comparisons to the space of
capabilities....dictates that life expectancy, the length of time over which
an individual has the capability of enjoying such things as he or she values,
should be included in the calculation of their well being. [27] |
|
blindingly obvious [28] mortality paradox [29] Any outcome measure that does not take
deaths into account can give paradoxical results. [30]
|
There is a glaring paradox in all commonly used measures of
poverty.[31] |
|
fact of mathematics...if
the poorest person in a group dies....per capita income of living people is
higher [32] |
technical-mathematical reason....removing the poorest person
from the count lowers poverty as measured. [33] |
|
not primarily an empirical issue
[34] This raises a conceptual issue [35] |
the
issue raised here is not empirical. It
is conceptual. [36] |
|
The logical fallacy that in poor countries (populations with
unknown differential mortality rates) “average income went up” is the same as
“people’s incomes went up” may turn out to be one of the most disastrous
logical mistakes in recent history. [37] |
It cannot be right in
concept that differentially higher mortality rates among the poor serve to
reduce poverty. [38] |
|
I
am working on a way of taking mortality into account when measuring welfare
outcomes for a population over time. [39] |
The
purpose of our paper is to derive instructive analytics on how to account for
differentials in demographic variables, and in particular mortality, when
performing welfare comparisons over time
[40] |
|
no outcome measure is
humane unless it takes into account what happened to people who started the
period but didn't make it to the end. [41] |
It
would be a monstrous assault on our fundamental intuitions if [AIDS deaths] were not recorded on the
negative side of the ledger... [42] |
|
all measures (proportion under
$1, poverty gap etc) look better [if
enough of the poorest die] [43] |
the most commonly used family of
poverty measures... would decrease if the poorest person died as a result of
poverty. [44] |
|
average income of those
alive will be exactly the same as if he had survived and raised his income to
the average [45] |
In principle we should project
what they would have had if they had been alive. [46] |
|
It
is standard practice in economics to use changes over time in
indices of average income among living populations. [47] |
The
standard approaches to poverty measurement look at a snapshot of alive
individuals [48] |
|
for the purpose of clarity the language used is simplified somewhat. No-one’s life is infinite, so whereas I talk about people being counted in the index at different times, in real life people may be replaced by their descendants. Also, a real population may grow, whereas I talk as if it doesn’t. [49] |
conceptual clarity can be gained by considering a situation in
which....each individual at income level Yi lives for li periods, after which
time he or she is replaced by exactly one individual. [50] |
|
no-one
knows how much [poverty reduction statistics are] due to raised income, and
how much to excess mortality among the poor due to AIDS, hunger or bad
government. [51] |
relatively little is known about....what exactly
it is that we are buying into when we accept [the goal of] halving income
poverty [52] |
|
if the poorest die, the
income figures look better. [53] |
|
|
if the poorest die, the income figures look better. [55] |
when a poor person becomes poorer the headcount index of poverty
will not increase (indeed, if the person dies, the index will fall!). [56] |
|
differential survival rates according
to wealth are clearly not constant over time, between policies, over the
incidence of disease, or between countries.
In fact, nor are they perfectly constant in rich countries... These
points on differential survival rates are in no need of empirical proof [57] |
higher
mortality rates and lower life expectancies among the poor are an established
empirical regularity the world over. [58] |
|
In Ravi Kanbur’s paper on disaggreements he points out the
mismatch between people’s perception of poverty (no improvement) and the
headline figures produced by macroeconomists. One reason for this could be
the disparity in mortality rates between the middle class and the poor [59] To study relevant data from the
past, then, it is necessary to look at what happened to everyone in the group
over the period. If they died, then we need to ask what happened to
their welfare before they died, and why they died. My own view is that any question
of what is good for the poor, or how the poor are faring, has to take
mortality during the period into account - otherwise the
statistics are vulnerable to this effect, in addition to discounting the
human cost (or saving) and ignoring what happened to the welfare of those who
died before the end of the period. [60] |
I think you’ve put your finger on something very important. [61] |
|
if the poorest die at a
disproportionate rate... average income of those living will automatically be
higher [62] |
In the presence of premature mortality for the poorer sections
of the population, standard snapshot poverty measures will show a
decrease. [63] |
|
Any measure of my
well-being has to take into account my likelihood of dying.... Mortality
rates reveal degrees of risk for individuals in the group. If the age-specific death rate curve
improves nationally, but gets worse for the poor, then poverty has increased. If the curve gets better nationally and
improves at a slower rate for the poor, then inequality has
increased....Measures of “income inequality” don’t apply where mortality rate
differences are so unequal. [64] |
I
tentatively suggest three candidates for topics or questions in the distributional
area that may well spark the conceptual excitement of three decades ago —
measurement of poverty with differential mortality rates.... [65] |
|
economists ask, for example, “was x or y good for the poor?” on the basis of average income for the bottom quintile of those living in a country at various times...[66] The mathematical relationship between a rise in average income among the living and a rise in income for the real population at the start is completely unknown, without a large amount of other data. [67] |
Current
conceptualizations of poverty measurement focus (somewhat unthinkingly) on
those currently alive...[68] |
|
If there is a possibility that statistics that look good are reflecting outcomes we consider bad, then a fundamental question arises as to the burden of proof, and another about what concepts and terminology describe accurately enough what is going on. [69] |
|
|
account needs to be taken
of those within the relevant group who did not achieve the target, whether
through death or any other path. [70]
|
We call the set of individuals who enter into the measurement
poverty, whether they are alive or dead, the relevant set of
individuals. [71] |
|
If economists don't tackle this, they
will end up saying that genocide of the poor is good for the poor...[72] |
|
|
Other things being equal, if someone
on below-per-capita income, at any age, survives a situation where they were
about to die, GDP per capita falls.
Amazingly, economics has never
taken this on board, even though it applies in rich countries as well as
poor. [73] |
|
|
Cross-sectional data
are static data from whoever is in the population at the time. Per capita statistics are on the basis of
cross-sectional data. They ignore
whether the people are adults or children, and the paradoxical effects of
longevity. [74] |
|
[1] Email to Amartya Sen. 19 July 2000. Automated reply received.
[2] Reforming poverty alleviation policies. Email to Jonathan Morduch, Princeton. 3 August 2000.
[3] Jonathan Morduch. Concepts of poverty. Draft chapter for UN Handbook on poverty measurement. 2005. http://unstats.un.org/unsd/methods/poverty/2.%20Chapter%202%20(15-06).doc
[4] Possibilities. Email to Simon Maxwell, Overseas Development
Institute, London, 27 July 2000.
[5] Reforming poverty alleviation policies. Email to Jonathan Morduch, Princeton 3 August 2000
[6] Angus Deaton, Princeton. Counting the world’s poor: problems and
possible solutions. August, 2000
Revised, December 2000.
[7] Possibilities. Email to Simon Maxwell, Overseas Development
Institute, London, 27 July 2000.
[8] Simon Maxwell, 21 September 2000.
By email.
[9] Erik Thorbecke, 2003
Conceptual and Measurement Issues in Poverty Analysis Paper prepared for United Nations University
-WIDER Conference on Inequality, Poverty and Human Well-Being, Helsinki. 30-31 May 2003.
[10] Document 16 October 2001. Written as reminder of contents of email sent to Andrew Smith MP 15 October 2001. Original forwarded by Mr Smith to Clare Short MP, Secretary of State for International Development and Britain’s Governor of the World Bank.
[11] Kanbur 2002 (as above).
[12] Document
16 October 2001. Written as record of
email sent to Andrew Smith MP 15 October 2001.
Email forwarded to Clare Short MP, Secretary of State for International
Development and UK Governor of the World Bank.
[13] Kanbur and Mukherjee 2003.
[14] Finance ministers, truth and logic. Part 1 of a report on the logic of economics. Submitted to ELDIS, University of Sussex; sent to Simon Maxwell, Overseas Development Institute; Alex Wilks, Bretton Woods Project; Professor Uskali Maki, Erasmus University, Rotterdam. Draft 26 January 2003.
[15] Note 28 June 2000. Point raised at DFID Development Forum, Church House, London, 5 July 2000 during Trade and Poverty workshop.
[16] David Hulme (Manchester) and Andrew Shepherd (Overseas Development Institute). Conceptualizing chronic poverty. World Development, March 2003.
[17] Millennium Goals and poverty indicators. Email to Tony Baldry MP, Chairman of the International Development Select Committee of the UK House of Commons. 25 June 2003
[18] Finance ministers, truth and logic. 26 January 2003 (as above).
[19] Email to Professor Anthony Giddens, London School of Economics. 19 August 2002.
[20] Jonathan Morduch, Concepts of Poverty. Draft chapter for UN Handbook on poverty measurement. 2005. http://unstats.un.org/unsd/methods/poverty/2.%20Chapter%202%20(15-06).doc
[21] Hilary
Benn MP, 2003. Letter to Tony Baldry MP
concerning widespread errors noted by Matt Berkley. Mr Benn, then Minister of State for
International Development, wrote “I am responding as monitoring of progress
towards the millennium development goals falls in my area of
responsibility.”. He subsequently became
Britain’s Governor of the World Bank.
[22] Gordon Brown MP, Britain’s Alternate Governor of the World Bank, November 2003 after examining a one-page note pointing out this error.
[23] Re:
Three questions on utility and life length.
Usenet post to sci.econ Jul 20,
2003 http://www.talkaboutinvestments.com/group/sci.econ/messages/188873.html
[24] Gordon
Anderson, 2004. Life expectancy and
economic welfare: the example of Africa in the 1990's. University of Toronto. November 2004
[25] Note 25 June 2000
[26] Re: Mistakes in macroeconomics?
Usenet post to newsgroup: sci.econ
17 July 2003
[27] Gordon
Anderson 2004 (as above).
[28] Possibilities. Email to Simon Maxwell, Overseas Development Institute, London, 27 July 2000.
[29] Draft letter to James Wolfensohn. Document saved 18 April 2001.
[30] Draft 9 April 2001. Same
day: Explained problem to Brian Hammond,
in charge of development statistics at OECD.
[31] Ravi Kanbur and Diganta Mukherjee 2003. Premature mortality and poverty measurement.
[32] Notes “Raw material for Wolfensohn letter” 11 April 2001
[33] Ravi Kanbur 2002 (as above).
[34] Draft letter to Ravi Kanbur between
conversations of 18 April and 2 May 2001.
[35] Draft letter to James Wolfensohn 18 April 2001 apparently saved
just prior to first conversation with Professor Kanbur.
[36] Ravi Kanbur 2002 (as above).
[37] Document 16 October 2001. Written as reminder of contents of email sent to Andrew Smith MP 15 October 2001. Original forwarded by Mr Smith to Clare Short MP, Secretary of State for International Development and Britain’s Governor of the World Bank.
[38] Kanbur and Mukherjee 2003 (as above).
[39] Life
and death. Email to Maria
Bonilla-Chacin, Johns Hopkins University, co-author of “Life and death among
the poorest”. 3 April 2001
[40] Denis
Cogneau and Michael Grimm 2004. The measurement of poverty dynamics when
mortality is correlated with income - theory, concept and empirical
implementation. First complete
draft June 1 2004
[41] Economics of survival.
Email to Jeffrey Sachs. 11 April
2001
[42] Ravi Kanbur 2002 Conceptual challenges in poverty and inequality: one development economist's perspective.
[43] Notes 10 April 2001
[44] Ravi Kanbur 2002 (as above).
[45] Economics of survival.
Email to Jeffrey Sachs 11 April
2001
[46] Kanbur and Mukherjee 2003 (as above)
[47] A modest proposal: Economics and variable mortality. A short note on set theory and inference from changes over time in groups with variable populations to real populations Draft 26 April 2001
[48] Kanbur and Mukherjee 2003 (as above)
[49] Welfare indices of the living depend on who survives. Draft 28 April 2001
[50] Kanbur and Mukherjee 2003 (as above)
[51] Document 16 October 2001.
Written as reminder of contents of email sent to Andrew Smith MP 15
October 2001. Original forwarded by Mr
Smith to Clare Short MP, Secretary of State for International Development and
Britain’s Governor of the World Bank.
[52] Ravi Kanbur. Growth, inequality and poverty: some hard questions. January 2004
[53] Reforming poverty alleviation policies. Email to Jonathan Morduch, Princeton. 3 August 2000.
[54] Growth may be good for the poor-- but are IMF and World Bank policies
good for growth? A closer look at the
World Bank's most recent defense of Its policies. Mark Weisbrot, Dean Baker, Robert Naiman,
and Gila Neta. Center for Economic and
Policy Research, Washington D.C.
Draft released August 7, 2000
[55] Reforming poverty alleviation policies. Email to Jonathan Morduch, Princeton 3 August 2000
[56] Martin Ravallion. Issues
in Measuring and Modeling Poverty, Policy Research Working Paper 1615, Policy Research Department, Poverty and Human Resources Division, The World
Bank, June 1996. Page 2.
[57] Welfare indices countires (sic).
Draft 25 April 2001
[58] Kanbur and Mukherjee 2003 (as above).
[59] Progress, poverty and survival.
Helping people out of poverty: A short note on measuring progress. Draft 9 April 2001
[60] Draft letter to James Wolfensohn. 18 April 2001
[61] Ravi Kanbur, 18 April 2001. Telephone.
[62] Economics and mortality rates.
Email to Kenneth Arrow, Stanford.
20 April 2001
[63] Kanbur and Mukherjee 2003 (as above).
[64] Progress, poverty and survival. Draft 9 April 2001. Same day: telephone conversation with Brian Hammond, in charge of development statistics for OECD in Paris.
[65] Kanbur 2002 (as above).
[66] Economics of mortality rates.
Email to Kenneth Arrow, Stanford
20 April 2001
[67] Welfare indices countiries. Draft 25 April 2001.
[68] Kanbur 2002.
[69] Social science and survival . Draft 2 May 2001. Same day: second conversation with Professor Kanbur.
[70] Economics of survival. Email to Professor Jeffrey Sachs, Harvard. 11
April 2001.
Also on that day: telephone discussion with Eric Swanson, head of World
Development Indicators project at World Bank. ?40 minutes?
[71] Kanbur and Mukherjee 2003 (as above).
[73] Finance ministers, truth and
logic. 26 January 2003 (as above).
[74] Finance ministers, truth and logic. 26 January 2003 (as above).