View Full Version : Afghan Population

Steve the Planner
10-28-2009, 01:41 AM
OK. So, I am getting tired of reading press coverage of the either 26 million, 30 million, or 33 million Afghans to which someone is trying to apply one metric or another to.

I looked up the World Gazeteer which shows a 2009 population estimate of 24, 316,392. There's one source, and world gazeteer also has a breakdown by major city for anyone who is interested which provides the 1979 Census and 2006 and 2009 Estimates.


Another source is the UN. UN sites are pretty similar---around 24 to 26 million projected as of 2008 or so.

UNDP's May 2009 Assessment Report indicates: 26,813,057 (July 2001 Estimate)--- Yes, that's what it says!


Who indicates 26,088,000 as the 2008 Who estimate based on 2006 figures.

UNHCR's 2008-2009 Returnee Report indicates 23,511,400 population plus 4,351,307 returnees. 27,8862,707 total?

Interesting that the refugee report shows approx. 1 million Afghan refugees in Iran, and almost 2 million in Pakistan. Another 500,000 or so are distributed worldwide. Their reports also include maps with pop estimates by region, and there are multiple estimated pops based on return scenarios. (Lots of data)


One UN Habitat site shows the Total Pop as 22 million as of 2001 with percent urban as 22%, and urban and slum annual growth rates of 6%.

The summary basis from Habitat is:

"The steady inflow of returnees has further exacerbated the problem of inadequate housing and infrastructure, particularly in larger cities like Kabul, Jalalabad, Mazar and Kandahar. There are reasons to believe that, as peace returns to the country, its urban population will increase as a much faster rate than its rural population. According to UN estimates, from 2000 to 2015, the national population is expected to increase by 14 million to reach a total of about 37 million; more than half of this growth will be in urban areas. Given the country’s strategic location for regional trade and transit, the entrepreneurial nature of its people and the limited capacity of rural areas to absorb further population growth, Afghanistan is destined to urbanised rapidly. In fact it is one of the most rapidly urbanising countries in the sub-region.

Urbanisation will pose both a challenge to, and an opportunity for the development of the country. The way the current pressures on its urban areas are managed will influence the success of its development endeavours."


Another data point is the Afghan Ministry of Urban Development. I downloaded an April 2009 Report from the Minister to UN Habitat Conference (Nairobi).

Although no overall pop stats are provided in this report, Minister/Engineer Pashtun reports that 5 million refugees have returned in the last few years with more than 70% going to urban centers. Additionally, there is a 5-6% annual rural to urban migration rate.

For Kabul, with 180,000 households, he reports more than 67% of the population of Kabul City lives in informal settlements, and 50% in the urban areas as a whole, with 1.67 families per household.

If I ever get the time, I might sit down and rationalize some of this, but one thing I have not seen is any report that suggests the 30 million or 33 million routinely cited in supposedly authoritative press sources and reports.

Anybody got a any better sources?


10-28-2009, 01:19 PM
I think you've hit upon all the major estimates. I will have to do some research, but if I remember correctly, the 30 million plus estimates were simply based on a higher populationg growth rate extrapolated from 1970's population data.

Steve the Planner
10-28-2009, 01:43 PM
Estimate vs. Projection.

A projection takes two prior points and projects the line into the future.

An estimate considers the projection, but also takes into account expected non-linear factors.

In these conflict and post-conflict zones, the expectation of non-linear results is high, but it depends on how good the analyst, data, or sources are.

In Iraq, the two prior census components allowed for a projection (1987 vs.1997), but the Min of Planning actually used estimates derived from their local DGs. In some areas, they were remarkably good, taking into account real-time factors such as IDP displacements, etc... The Food Ration registration numbers were, apparently, often used by the DGs, together with decades of local knowledge about every village and family, to get pretty close (95% +/-)

Trying to figure out who is using which approach in Afghanistan, and what they are factoring in. Wonder what proxy sources (like Food Rations in Iraq) that might exist for disaggregated projections, at least down to district/major city levels.

Background issues for me are the split between urban/rural and the inherent socio-economic factors represented therein.

Americans always say Afghanistan is just rural as can be while international sources point to the north and west as rapidly urbanizing, consistent with multi-national regional rural-to-urban patterns. I suspect that each view is based on where you stand, and what you see if front of you, but, more likely, there are two materially different patterns in different areas.

Be nice to know, with a little more acuity, how many districts are urban, and where they are---in real-time, and not based on outdated projections.

Of course, the momentum patterns are the big issue. What is the dynamics.

Sounds like if 22 million were 22% urban in 2001(5 mil?), five million refugees returned with 70% going to urban areas (3.5 mil), and background rural-to-urban patterns increase urbanization by 5-6% per year for eight years, the final calculations, assuming a 26 million total pop, may be not unsurprisingly like the Minister reported to the UN: Urbanization is a major problem for Afghanistan.

The internal pressure alone from this mass land use change is just huge, let alone the secondary political and humanitarian factors.

But, if we are going to re-posture to focus on Afghanistan as a whole, rather than just the rural Pashtun areas, it would be great to know what that means.

Maybe we can piece some of it together?


10-28-2009, 03:11 PM
A lovely stats related quote I picked up somewhere states, "You can torture statistics enough to admit to anything."

Population figures in Afghanistan are ripe for torture since there are no reliable or consistent figures to refer to. CIA World Factbook uses the higher end figure of approx 33M versus the Afghan Central Statistics Office (CSO) of approx 27M (for 2009) which are all growth based projections from 1979 census data. The UN data office (statistics division) uses a figure of 27M for 2008 and promotes a population growth rate of 3.9% (or 5.4% urban/ 3.4% rural) which would place the settled population close to 29M today. If this means IDPs/refugees are not included, then we could be at or above the 33M figure. Rural population is quoted at 27% but msot of Afghanistan is non-arable making even the rural population fairly concentrated to rivers and valleys throughout the regions.

Urban centres are under tremendous pressures due to in-migration within the 12 major cities which is influenced by operations in the rural areas and the draw of the PRTs and major NATO bases.

The CSO has been planning to conduct a new census for at least the last 3 years and currrent planning is focussed for this to occur in 2010. Until then, there are no reliable figures nor reliable trends to rely upon.

Steve the Planner
10-28-2009, 03:29 PM

Thanks for the contribution.

I'll scratch my head on this piece as soon as I get a chance.


Steve the Planner
10-29-2009, 02:27 AM

Punched through to the Central Statistics Office site. Great site.


23,993,500 is the 2009 Estimate without External Displacements. Breakdowns by province are provided from 2002 to 2009.

UN usually has a pretty good survey program for Displaced People to understand which refugees will and will not return, and to where. Question is whether the 4 million or so refugees would come back as soon as any stability level was attained.

CSO apparently uses the same UN training and methodologies as Iraq's CoSIT staff. Pretty good stuff.

The Ottoman Empire was grounded in per unit taxes (livestock, agricultrual production, etc...) so Iraq's government had always been good at counting everything. Their only limitations were in mobility, new technology, and destroyed historical records.

Still, conducting a census is always a serious undertaking, but especially in a war zone. NGA and Iraq's Defense Mapping Agency were helping the Ministry of Planning with the census, but the shear number of local maps (for census workers) was in the thousands. They are shooting for next year (fingers crossed).

As for rural/urban, UN Habitat is significantly focused on "informal settlements" as well as formal ones. Perhaps some folks on the ground mis-understand the 27% rural figure to mean something else.

Not surprising in a small farm structure, as with Iraq, for every family to have a plot outside of town (and thus, be farmers), but live and work in the nearby "town," however that may be defined.

The big elephant in the room is the 60% under age 25---the "ticking time bomb" if they cannot find legit work.


Steve the Planner
10-30-2009, 12:58 PM

I found it interesting that you should note that PRTs, ISAF, NATO themselves are a significant land use/population settlement pattern driver---as much as an attractor as a refugee driver (flight from conflict):

"Urban centres are under tremendous pressures due to in-migration within the 12 major cities which is influenced by operations in the rural areas and the draw of the PRTs and major NATO bases."

Very insightful, and important to remember.


Steve the Planner
10-30-2009, 01:51 PM
Gradually, we are making some headway.

With apologies, the CIA Factbook has now revised it's Afghan Population figures from 33 million down to 28,396,000.


If they can do that, perhaps they can update those really bad Iraqi provincial boundary maps.

Who says an old dog can't eventually learn some tricks?


Steve the Planner
11-02-2009, 05:28 PM
Link to the "In the Loop" column of the Washington Post.

Al Kamen's write-up- "All those Afghans, give or take a few million"- sums up the latest significantly revised population figures. Down from 33.6 million to 28,396,000.


CIA had, apparently, relied on the US Census Bureau for its Afghan estimates.

I'm still looking at the UN figures in the 26.5 million range, so another 2 million remains in limbo, but the new figure is better than the old one.

Now, how are the changes allocated to the provinces, districts, metrics, etc...?

PS- Obviously, somebody outside of our circle actually reads this stuff.

Steve the Planner
11-05-2009, 06:28 AM
US Contractors in Afghanistan

A June 2009 Pentagon report indicated the number of US contractors in Afghanistan is 74,000, and, no doubt, rising. So, 1.1 contractors for each soldier?

Expat Experts

One report out of Rory Stewart's Harvard shop indicated that there were 5,000 expat experts in Kabul. I haven't found a source for that, but with UN's recent announcement that 900 UN workers are being removed, I assume that number will take a big hit.

I wonder how many expats you have to accumulate in a capital city before the shear weight of their bar and restaurant spending creates a nationwide economic boom?

Big controversy right now in Baghdad is why KBR hasn't reduced its staff as troop counts go down. I'm not an apologist for KBR, but if they are still maintaining everything they were before on all the big FOBs, plus working on shipping us out of there, is it illogical that their numbers might spike for a little while.

Regardless, as of October 30, the Green Zone is officially "dry" as a result of Iraqi orders. With the IZ liquor store shutting down and restaurants halting alcohol sales, it seems likely that Green Zone civilian staffing will plunge dramatically.

Steve the Planner
11-21-2009, 05:09 AM
Afghan Central Statistics Org was off-line for a little while.

Latest 2009 Pop Estimate: 23,993,500.