As Sri Lankan income increases, people are spending on taste (‘other’, condiments) rather than calories or protein.
The average Sri Lankan meal is rice, vegetables, and maybe fish, cooked with a lot of chili and spice. This is common knowledge, but it’s also what I found in the latest Household Income And Expenditure Survey.
This survey of 22,581 people has data down to what type of rice (Kekulu, mainly), what type of vegetables (beans and brinjal), what variety of fish (small ones, I assume hal maso) and what condiments (Bombay onions, green chilis).
What Sri Lankans Are Eating
What Sri Lankans are eating – still mainly rice and coconut milk curries*
Sri Lankans are eating mainly rice and coconut milk infused vegetable curries. Note that quantities for vegetables and fruits are not included here, because the data only tells you how many bananas a person eats per day (0.4) and not how many grams of banana. This is a pretty big omission in their results which I don’t have the energy to recalculate from the data.
Just eyeballing it, people are eating a lot of Bombay onions, a lot of dhal (a pulse, really), a lot of beans, brinjal (eggplant) and pumpkin, in that order. This is literally what they have at the kade down the street and also what I had for lunch (first three). Maldive fish is used liberally as a spice and people eat fish (hal maso, teppili, or salaya) if they can get it. This is basically how I eat.
What Sri Lankans Are Spending On
Sri Lankan expenditures on different food items. ‘Other’ is now more than rice.
To observe one top-level thing, while what Sri Lankans eat hasn’t changed much, how it tastes has. With more income, Sri Lankans are opting not for more food (actually less) but better tasting food. That means they’re spending more on condiments (spices) and ‘other’ (sauces, tinned or packed stuff).
In terms of this study, other refers to sauce (ketchup), marmite, soup cubes, pickles, chutneys, canned stuff, tinned stuff and – I assume – prepared food bought from outside. I assume the latter because while the average Sri Lankan spends about Rs. 90/day on rice, they spend about Rs. 50 on food from outside (bread, string hoppers, etc). The numbers for other don’t work if it’s just ketchup, but they do if you include eating out.
ASIDE: The average person eats 16.19 string hoppers per month, or about 1/2 a string hopper per day. I recently crashed at the Tissawewa Grand in Anuradhapura where they gave us literally 90 string hoppers for three people, or twice the average monthly dosage. I ate maybe 10.
Taste Not Calories
Basically, as Sri Lankans income increases they’re not spending it on more calories or even more protein, they spending it on spices and manufactured goods that make food taste better. Which makes sense on a kade level, but is still baffling to mainstream economics and development theory, which sees third worlders as hungry people trying to maximize calories.
People like Abhijit Banerjee are pushing for a more realistic (and less ‘rational’ model), but it’s still a minority view that doesn’t shape much global policy. What I find surprising is you can actually figure all of this out if you just visit the kade down my street. I buy rice, sugar, potatoes, tomatoes and, if I have money left over, manufactured crap. As he and Esther Duflo wrote:
Studies have shown that when very poor people get a chance to spend a little bit more on food, they don’t put everything into getting more calories. Instead, they buy better-tasting, more expensive calories. (Foreign Policy)
Averages
Anyways, the average household spends about Rs. 13,000 on food per month. Unless you go to Arpico, where you can spend that much and not really feel like you’ve acquired anything. On average, this is 42% of a person’s total household expenditure. It varies, in Colombo people spend about 34%, in Jaffna about 62%.
In rural areas, they spend on kekulu rice and cereals, The highest spend is 29% on cereals in Nuwara Eliya. In Colombo people spend a lot on prepared foods (15.5%).
They type of rice you eat is pretty clearly tracked to income
Rice is interesting because it breaks pretty clearly along class lines. In the estates they eat Nadu rice, in the rural areas its Kekula and in urban areas its less overall, and more Samba.
So, that’s what the report itself has to say about Sri Lankan food consumption.
An Average Meal
The tables breaks stuff down to what vegetable, to what oil, so I can hazard a guess as to what an average Sri Lankan meal would be:
Kekulu red rice, with dhal, a potato curry, beans and brinjal. All of the curries would have fresh coconut milk, be cooked in coconut oil and be liberally spiced and flavored with Maldive Fish flakes. They’d be seasoned with whole green chilis and chili powder. Every now and then you’d get a fish curry, most often hal maso (sprats). 12 days out of 30, you would be offered a banana. You would also get tea, with a shitload of sugar, especially if you’re a guest.
The original data is from this PDF, the Household Income And Expenditure Survey (HIES) from the Ministry of Finance, Sri Lanka. PDFs are evil and I dumped some of the data into a Google Doc, but that’s a mess and I barely understand it myself. It does, however, include a nice GapMinder style interactive chart.
Is there any nutritional benefit to coconut milk? Is coconut oil still the most commonly used oil? I personally dont like coconut oil, it feels really heavy. But I absolutely love coconut milk and I’d love to find out that its packed full of vitamins, or something.
Not sure about the milk, but It’s probably not that harmful either. Saw an article in yesterday’s Times, about new research concluding that coconut oil isn’t as bad for you as they thought it was. Also said the traditional method of extraction by boiling the milk and skimming the oil that comes out, is as good at maintaining nutritional value as extracting virgin oil, as opposed to the method commonly used at a mill. Very comforting news for us Sri Lankans. Nutritional value here means anti-oxidant properties. Apparently the anti-oxidants present in c’nut oil are also highly thermally stable and tend to dissolve more in the oil simmering at high temperatures, which makes the traditional method a good one, and also makes fried/sauted food less harmful than we thought.
I have an aunt who rarely buys c’nut oil but makes a huge batch of kalu dodol regularly and uses the oil that is skimmed from it for cooking. Only, the process consumes time, energy and fuel (requires continuous stirring for hours over a hot fire). I have this manually operated gadget ( Rs. 3500/- from NERD ) which extracts virgin oil by squeezing dried grated c’nut, and the refuse can be sqeezed with a little water in the same apparatus to extract low fat milk for cooking. Not very economical as all the oil doesn’t get extracted, but it works out OK for me as we have a few coconut trees.
Thanks for this article. Well research and very informative!