Pakistan’s wheat demand, in the words of Winston Churchill, “is a riddle wrapped in a mystery”. But equally difficult is to decode the magic number of 124kg per capita, bandied around as the reference figure in official as well as non-official sources alike.

What is the ultimate source of this magic number? One easy way to arrive at it is to divide the average annual supply of wheat (opening stocks less of external trade plus season’s output) and divide it with the population figure, and one may arrive in the comfortable range of 122-126kg, varying slightly from year to year.

But is it not surprising that Pakistan’s high per capita consumption of wheat is unmatched by any comparable nation, whether within the neighbourhood, or one with similar income-levels? Granted that in most East Asian countries, rice and not wheat, is the staple crop, but the per capita boasted by Pakistan does not match up even if one were to turn westwards to the brotherly neighbourhood of Islam. Moreover, does Pakistan’s wheat crop not suffer from any wastage, spoilage, or loss post-harvest, storage, transportation and processing?

The indomitable Household Integrated Economic Survey by PBS discloses average monthly domestic expenditure on cereals, specifically wheat flour, which can then be used to calculate volume of wheat flour consumed by households.

Using average per kg wheat flour price during FY16 – the year of latest PSLM survey –volume consumed per household and per person has been estimated in Figure 1, which arrives at per capita wheat consumption of no more than 85kg. That’s just 70 percent of total domestic reference number of 124kg per capita, leaving 7 million tons for industrial consumption and spoilage.

But does the HIES figure itself hold up? The key number (or assumption) in the HIES survey is the 12.55 percent of monthly expenditure budgeted for wheat flour, which there exists no independent way of verifying.

Surprisingly enough though, same volume and value estimate as HIES can be arrived at using a basic back of the envelope market sizing technique: average daily per capita consumption of wheat flour of 240gm. Underlying assumption? Four rotis per person and 60 grams per roti as reference weight. The rest is just multiplication and division using total population and nominal prices, and the reader may arrive at 88kg per capita and monthly household expenditure of Rs1,727 per household on wheat flour, compared to HIES-based figures of 85kg and Rs1,682 per household – in the same vicinity as HIES.

This is not to imply that the HIES numbers use a similar reverse-engineering technique. However, even intuitively, availability of four rotis for entire population all year round does appear very generous, as it neither takes into account the prevalence of hunger and undernourishment at close to one-fifth of population, and the more intuitive weaknesses in the assumption – which wheat-based caloric consumption varies highly across ages and genders.

Yet, even if one were to rely on the HIES-based estimates, the gap between annual demand and supply of 30 percent remains inexplicable. A generic assumption of 10 percent on farm use as feedstock – common for most cereal and grain crops – may still knock off some 2.5 million tons, with twice as much still to go.

Is five million ton too high for commercial/industrial consumption, or too low? In the absence of broad-based studies of cross-sectoral and demographic consumption trends, all calculations are at best mere guesstimates. Yet, neither the government nor industry association appear too keen. Donors: pay attention!