Superficially, the origin of the poverty line may not appear to be a problematic issue, as
long as it seems reasonable and acceptable, since it is just a means of allowing one to focus
on a particular part of the welfare distribution. However, the implications of choosing a
particular method can be far-reaching, since poverty measurement is concerned mainly
with making comparisons of poverty between geographical areas, groups in society and
over time. Using a relative poverty line or a different absolute poverty line per context may
mean that one considers a different norm for different countries or groups. For example,
an acceptable norm in a rich society may imply having the means to pay for rather fancy,
branded clothes, while in a poor society simple clothing may be sufficient. Arguments can
be made for such a method and when looking within the group or society in question, this
does not have to be a problem. But when making comparisons between societies, for
example to prioritise spending between countries, allowing for these different ‘needs’ will
have a different implication from using basic clothing for all societies. Similarly, should
one compensate one social group for its culturally determined but more expensive tastes
when constructing a poverty line? Or, when evaluating the change in poverty over time,
an ‘absolute’ fixed basket of commodities will have different implications from using an
updated relative poverty line, linked to the mean or median welfare outcome in each
period, as used in the European Union. For example, adjusting the poverty line to the
mean may seem attractive for a broadly growing economy – resulting in lower poverty
declines than may be implied by using a fixed line. But when a big crisis hits an economy,
halving its GDP, without any change in income inequality then the poverty line is likely
to halve as well and no increase in poverty will be observed as if this has no implications
for poverty. The alternative, such as using a fixed poverty line reflecting the purchasing
power implied by one US dollar in a particular year, as used in the Millennium
Development Goals (MDGs), provides then a relatively transparent tool for comparison
between countries, and possibly allocating funding between countries, even though as a
means for conducting a substantive investigation of poverty within a particular country
it may seem wanting (see Millennium Development Goals, this volume). These examples
suggest that poverty lines should be chosen to serve best the nature of the poverty comparison
one wants to conduct over time, between groups or between countries and
regions. Each choice will partly predetermine the findings, and any poverty analysis
should be explicit about the reasons for its choice.