Transitioning to the New 2011 PPP Lines >
As you may know, in October 2015, the World Bank released the new International Poverty Line. The new line boosts the 2005 level, $1.25 based on 2011 local prices converted to US dollars, up to $1.90/day. While it seems like a simple adjustment, measuring poverty across times and countries is an inherently fuzzy process. The purpose of this second installment of the PPI Practitioner Guidance Series (read the first installment here) is to explain why these two lines are not necessarily equivalent at the level of individual countries, and how PPI users who are currently using the 2005 PPP lines to measure poverty may transition to using the 2011 PPP lines.
How were the $1.25/day 2005 PPP and the $1.90/day 2011 PPP lines derived?
These detailed FAQs released by the World Bank explain how these lines are derived. The $1.25/day line was derived by taking an average (in 2005 PPP terms) of the National Poverty Lines of 15 of the poorest countries in the world. The World Bank regularly updates the International Poverty Line to reflect changes in cost of living over time. The $1.90/day line has been derived by averaging the National Poverty Lines of those same 15 countries (in 2011 PPP terms).
The World Bank chooses the same 15 countries to update the International Poverty Line to ensure that global poverty goals remain broadly consistent even after the revision. Thus, the World Bank estimates that in 2011, the share of the world’s population under the $1.25/day 2005 PPP line was 14.5%. For the same year, the global poverty rate under the $1.90/day 2011 PPP line was estimated at 14.2%.
For individual countries, are poverty rates below these two lines identical?
Not necessarily. To understand why, look at the chart below from the World Bank’s Policy Research Working Paper WPS 7432, Ferreira et al 2015: “A Global Count of the Extreme Poor in 2012: data issues, methodology and initial results.”1 This chart plots the 2011 PPP equivalent poverty line to the $1.25/day 2005 PPP line for 101 countries. As you can see, the line lies above or below the dashed line (which represents $1.90/day) for several countries. However, a population-weighted average of all of these lines would result in a figure of $1.93/day,2 very close to the dashed line.
Source: Ferreira et al 2015, World Bank Policy Research working paper; no. WPS 7432
Naturally, this has an implication on the poverty rate measured under the $1.25/day 2005 PPP and the $1.90/day 2011 PPP lines for individual countries. If the two lines are not equivalent, then the poverty rates measured against them will also not be identical.
Look at the chart below from the same World Bank Policy Research Working Paper WPS 7432, Ferreira et al 2015. This chart plots the latest available poverty headcount rate for the $1.25/day 2005 PPP line on the X-axis against the rate for the $1.90/day 2011 PPP line on the Y-axis for these same countries.3 If both rates were exactly equal, you would see the dots converge against the diagonal line in the chart. Dots above the line indicate that the poverty rate against the $1.90/day 2011 PPP line is higher than that measured against the $1.25/day 2005 PPP line for that country. Dots under the line indicate the opposite. Of course, the closer that the dot is to the line, the higher the degree of convergence between the two poverty rates.
What does this mean for you as a PPI user?
First, understand that the $1.25/day 2005 PPP and $1.90/day 2011 PPP lines are not necessarily equivalent for your country. It may be helpful to look at the values of these lines in local currency units, which is reported in the PPI Design Documentation released with every PPI. For example, in the case of Guatemala, for 2014, the local currency equivalent of the $1.25/day 2005 PPP line is 9.16 GTQ, while that of the $1.90/day 2011 PPP line is 8.35 GTQ.4 This may not seem intuitive to a user—$1.90, after all, seems like a “higher” line than $1.25—but do note that they are derived using different PPP factors and reflect the difference between changes in local currency compared to international inflation.
Second, this means that if you transition from the $1.25/day 2005 PPP to the $1.90/day 2011 PPP line, you cannot automatically assume that the poverty rate of your customer base will be the same under the two lines. The difference between the two poverty rates will vary by country.
So how do you reconcile these two lines? Simply treat them as two completely different lines. If you have previously set poverty outreach goals using the $1.25/day 2005 PPP line, measure change in poverty over time using the same line. Similarly, set separate goals and measure progress under the $1.90/day 2011 PPP line. Do not attempt to compare or combine poverty estimates obtained using these two lines. All PPIs released since 2016 are calibrated to both, 2005 and 2011 PPP lines. We will continue to support both lines for the next few years, allowing users time to make a complete transition to the 2011 PPP lines.
And finally, how do you set fresh poverty outreach goals using the 2011 PPP lines? We recently released a blog post offering advice on this. The second approach we outlined there would be helpful:
- Apply the PPI to your entire customer base or to a random and representative sample.
- Measure your poverty rate below the various 2011 PPP lines your country’s PPI5 is calibrated to. All PPIs released since May 2017 come with a simple Excel-based Data Analysis Tool which accepts the responses to the 10 PPI questions as input, and automatically computes PPI scores, poverty likelihoods and rates below all poverty lines (National, 2005, 2011, and Percentile) to which that PPI is calibrated.
- Use these results to determine future poverty outreach targets under the 2011 PPP lines.
For more advice or support as you navigate this process, please feel free to reach out to us at ppi@poverty-action.org.
[1] Ferreira, Francisco H. G.; Chen, Shaohua; Dabalen, Andrew L.; Dikhanov, Yuri M.; Hamadeh, Nada; Jolliffe, Dean Mitchell; Narayan, Ambar; Prydz, Espen Beer; Revenga, Ana L.; Sangraula, Prem; Serajuddin, Umar; Yoshida, Nobuo. 2015. A global count of the extreme poor in 2012 : data issues, methodology and initial results. Policy Research working paper; no. WPS 7432. Washington, D.C. : World Bank Group. http://documents.worldbank.org/curated/en/360021468187787070/A-global-count-of-the-extreme-poor-in-2012-data-issues-methodology-and-initial-results
[2] See Page 32 of same World Bank Policy Paper - “A Global Count of the Extreme Poor in 2012”
[3] Chart uses 2014 and 2015 versions of PovcalNet
[4] Refer to Table 1 on Page 147 of Guatemala’s PPI Design Documentation – available on our Guatemala PPI country page.
[5] All PPIs released from November 2017 onwards are calibrated to four international 2011 PPP lines: $1.00/day, $1.90/day (Extreme Poverty Line as defined by the World Bank), $3.20/day (lower middle-income International Poverty Line as defined by the World Bank), and $5.50/day (upper middle-income International Poverty Line as defined by the World Bank). Please see this World Bank blog for more on these international poverty lines. The $3.20/day and $5.50/day 2011 PPP lines were formally released by the World Bank in October 2017. Prior to this, select PPIs were calibrated to the $1.90/day and $3.10/day 2011 PPP lines. Ferreira et al 2015, World Bank Policy Research working paper; no. WPS 7432 also makes reference to the $3.10/day 2011 PPP line. This line was later updated by the World Bank to $3.20/day.
