I think I mentioned the potential factors for this in another thread:
https://defence.pk/threads/dhaka-un...ies-with-pakistan.413238/page-18#post-8148557
I don't know the exact sampling pattern of ICP when it comes to gathering the price level data for Bangladesh.
It is well known that PPP is somewhat artificially depressed in China because they have only allowed price level for their cities (as opposed to say India where much rural price level data are incorporated).
So maybe if the sampling is only within Dhaka and Chittagong, the PPP multiplier is also depressed because of it...but I am unsure because when I last read the ICP in depth, China was the only notable "exception" w.r.t sampling strategy. I mean it is done by BBS in the first place and then vetted by IMF through ADB etc.:
http://www.bbs.gov.bd/PageWebMenuContent.aspx?MenuKey=189
The main reasons (if its not the sampling pattern) I think may be a) access to natural resources (incl agri) b) dominance of RMG industry in Bangladesh exports/industry. c) land cost
i.e more import reliance and export price sensitivity...which do not imprint on PPP that much since the first takes away from GDP in solid US dollars (and hence impacts local price levels all across the production chain like a domino effect since you are paying the extra margins to an outside party for the base resource) and the 2nd is a large chunk of production that earns USD in the end and not local Taka within the economy (i.e the economies of scale have a greater impact externally rather than internally).
All in all there maybe should be some sort of composite measure since there are pros and cons for both PPP and nominal. But for now we will just have to quote them side by side and understand the limits of each.
It will be interesting exercise to know the cost of staple food items within Bangladesh and we can try compare some of the issues ourselves. I would imagine per unit area land value (general rural and outside suburb land) is also on average much higher in Bangladesh, this is a big big factor in PPP since access to cheap land has a huge multiplier effect in price levels of industrial goods and consumables since the land rent/cost must be recuperated through the pricing of the goods made on the land. So Bangladesh population density may play a big role in the end.
EDIT: Here is a good publication I just found:
https://www.imf.org/external/pubs/ft/scr/2010/cr1056.pdf
This picture presents the issue in a nutshell:
i.e the price level in Bdesh is increasing faster w.r.t per capita GDP compared to the average worldwide rate. The factors of production must be more dear in Bangladesh (raw inputs with the exception of labour)....and exposure to US dollar price level higher because of dependence on RMG export. As Bangladesh diversifies and bulks up (and does the harder economic chain development through human capital improvement), I would imagine it will smoothen out (i.e a large part of the price level increase basically got shoved into the beginning, so the benefits will come with time I think)