The reform of food subsidies under the Raisi administration
There is no shortage of grievances in Iran, so explaining the recent protests that have engulfed the country is not hard. The economy has stagnated for over a decade and inflation has been very high. But the timing of the protests suggests a connection between the protests and President Ebrahim Raisi’s decision to cut food subsidies. Some media reports speak of “bread protests”, but the price of the staple bread has not changed.
The events raise an important question regarding the protest: Are rising prices — for chicken, cooking oil, dairy and the like — cut into the food consumption of ordinary people? Will the program swell the ranks of the poor at the time when food budgets have been squeezed for years?
Last week, the Raisi government unveiled its first major economic policy initiative to reform food subsidies. Not a predictable move from a government that calls itself “revolutionary”, but it makes a lot of sense. Since 2018, each year some $10-15 billion of foreign currency has been sold to importers for food at 42,000 rials per USD, which is far below the free market and the semi-official rates in the so-called nima market (where licensed importers and exporters trade), about 5-6 times the subsidized rate. So, talk of removing the subsidy has been around for a while, and the recent global increase in food prices and anticipated shortages in coming months may have forced the government to act now.
Raisi’s program shares a key feature with the 2010 energy price reform of the former president Mahmoud Ahmadinejad, the forerunner revolutionary administration. Both programs combine price increases with cash transfers. Whereas the 2010 program focused on energy price reform, the aim of the current program is to remove the foreign currency subsidy for food imports. Raisi’s cash transfers are less generous than Ahmadinejad’s — which is a good thing — and is not universal. Individuals in the bottom 40% of the income distribution receive 4,000,000 rials each per month (less than one-fifth of the minimum wage) and about three-quarters of the 2010 transfers in real terms. Except for people in the top 10%, who receive nothing, the rest receive 3,000,000 rials each.
Despite similarities, the responses to the two programs in Iran’s city streets have been quite different. In 2010, people reacted to Ahmadinejad’s hefty energy price increases calmly, presumably because they figured they would come ahead in the exchange of energy subsidies for cash transfers, and most did, especially in the first 2-3 years.
Earlier this month, the Raisi administration started implementing its food subsidy reform program, depositing cash in people’s accounts while raising the prices of basic food items several fold. Protests broke out soon afterwards. An ill-fated gasoline price increase in 2019 ignited similar widespread urban protests, but then the government of president Hassan Rouhani, in keeping with his dim view of cash transfers to the poor, did not couple the price increase with compensating cash transfers. (It did a few days later but it was too late).
Given that Raisi’s subsidy reform offers generous compensation for higher food prices, why is it facing violent protests? Do protesters believe the compensation is inadequate, that the program will leave them poorer? Do they fear rising general inflation that will leave them poorer? Or, are they angry for any number of reasons?
In this post, I will take up the question regarding the impact of the program on living standards. Some simple arithmetics tells me that about half the people will not experience a decline.
The government has so far announced cash transfers for two month. Assuming that it will continue them for the rest of the year, simple calculations show that about half of all Iranians will come ahead. More precise estimates of the balance of higher food prices and cash transfers has to wait for the 2022 round of Household Expenditure and Income Survey (HEIS), which will be released a year from now, or more sophisticated modeling of demand for food. In the meantime, the 2020 round of the survey offers a rough guides as to who wins and who loses from Raisi’s program.
My rough estimates below are based on two further assumptions. First, assume that the distribution of income and expenditures this year remains the same as in 2020, and prices this year will continue to increase at the same pace as last year (40%). That means that the average price level will be twice as high as in 2020. This allows me to multiply all expenditures by a factor of 2 to obtain expenditures for 2022. I then add the cash transfer to the appropriate household expenditures in the relevant deciles and deduct the increase in food expenditures to arrive at the net transfer. Second, I assume that overall food expenditures will increase by 50%, a bit more than the general inflation. A more realistic simulation would use income and price elasticities to predict food expenditures, but those are beyond my limited aim here.
These assumptions lead to the conclusion that about half the population may be net winners. The table below present the distribution of transfers and per capita household expenditures in rials per day before and after the transfers. Net transfers suggest that the individuals in the first 4 deciles of per capita expenditures gain, the middle decile breaks even, while those in the top 5 deciles lose. The average loss is very small, about 20,000 rials per person per day.
| Decile | Net transfer | Food exp. pc | Total exp. pc | Exp. pc after transfers |
| 1 | 90551 | 85565 | 201268 | 292440 |
| 2 | 70942 | 124783 | 304858 | 375816 |
| 3 | 58037 | 150592 | 394710 | 452766 |
| 4 | 46146 | 174374 | 482025 | 528172 |
| 5 | 1515 | 196969 | 577253 | 578765 |
| 6 | -11400 | 222801 | 688138 | 676760 |
| 7 | -27934 | 255868 | 828886 | 800982 |
| 8 | -46595 | 293189 | 1018615 | 972043 |
| 9 | -83291 | 366583 | 1340216 | 1256933 |
| 10 | -275466 | 550932 | 2477534 | 2202089 |
| Average | -22360 | 248844 | 856265 | 834377 |
These rough calculations also show that the distribution of incomes improves — the Gini index decreases by 5 points, from 37.7 to 32.7, which is not nothing, and the poverty rate decreases substantially, from 13.6% to 6.1%.
As costly as the reforms have been in terms of social disruption, they do not fully remove the distortions caused by multiple exchange rates. Most of the wheat flour sold to traditional bakeries (khabbazi flour) will continue to be heavily subsided while those for other uses, such as for macaroni and baking sandwich rolls and sweets, will be sold at cost. The prices of these items have already jumped up, but ordinary bread is still selling as before, about 5 cents per 300-gram loaf. To limit demand for bread, the government plans to introduce smart cards that will allow each individual to purchase a fixed amount of bread at the current low price.
Of all food subsidies, those for bread are the most defensible. They are evenly distributed between the poor and the rich and reach the most vulnerable groups, like children, more effectively. However, given the ten-fold gap between the prices of flour for bread and other uses, incentives for illegal arbitrage are very high. There are already reports of bread bakeries selling their khabbazi flour on the black market and cutting their baking hours. This may cause long lines or a higher price for the staple breads, in which case consumers will bear a part of the cost of the dual price system.
Tyranny of numbers 
Hello Prof. Salehi,
I have a question which perhaps doesn’t have to do with the content of your current article directly.
What’s your view on Iranian governments’ (as they all unite on this point, reformist and principlist alike) strong emphasis on private sector and staying away from state owned investments both in production and also in import? So for example wouldn’t it be better to instead of giving dollars at discounted rates to private individuals to import wheat and make a profit which will obviously come out of the pocket of the consumers, that the government imports the wheat with the dollar that it has and sell it at a lower price (even at zero profit if necessary) in state-owned bakeries?
This was an example in giving privilege to the private sector in import but similar arguments can be made in case of production of certain commodities. Private sector by its very definition is after profit. For that they usually tend to put an emphasis on export whose market is often dominated by west and also vulnerable to sanctions.
Wouldn’t it make more sense that when it comes to certain commodities, that the state steps in for production? It would create jobs, and also it would open the path to produce certain commodities with an emphasis on meeting certain needs of the country. They have already had this approach to certain commodities required by military with success, what do you think of expanding that approach to production of some other strategic commodities badly need?
Salam, the prediction that inflation will reach close to 80% and above seems to be more realistic, especially if the JCPOA agreement is lost. see this one: asriran.com/003XXZ
Minimum wage this year is 57 milion Rial, as stated by Minister of Labour last March, thus 4 million Riam subsidy is no more than 7% of current minimum wage
Thank you for this information. 57 mil rial is high. If this takes effect, inflation would be more than the 40% that I assume for this year. I am comparing expenditures of the lower deciles in 2020 (1399) x 2 with the cash transfer, which comes to about 6 mil rials per month. 57 mil rials is nearly ten times that, One thing to note is that the min wage does not cover the majority of low wage workers.
Djavad jan, what was sent around was your blog on Ahmadinejad’s reform from 2011. I now see your new blog. Thanks, this is very timely and informative. I have 4 related questions/comments.
1. Is it possible to replicate this exercise using data at the time of Ahmadinejad’s energy reform to see how this approach’s prediction would compare what what happened? I ask this question because as it turned out ensuing inflation and fiscal problems (as you have noted elsewhere) among other things prevented a successful implementation of his reforms.
2. In light of Ahmadinejad’s experience, isn’t your assessment a bit optimistic? I think a crucial variable is overall inflation. Isn’t likely that it could well be higher than 40 percent? How is the government going to prevent the spread of inflation?
3. You note the government has for now promised 2 months of cash transfers but almost in passing assume this will continue for the rest of the year. Isn’t this a huge assumption? What if they do not? What if they do and the fiscal space doesn’t allow it? In both cases the outcomes would be quite different.
4. Finally, I think maybe it is for the above reasons that people in the streets have reacted the way they have. First, they remember Ahmadinejad’s reforms and second, they’re probably even more concerned about inflation than in the past.
Great questions. Here are my not-so-brief answers:
1. Yes, if you mean simulate the distribution with wild estimates of increased energy. It would be more positive because the poor spend much less on energy than they do on food. It would show they come well ahead, as they did. This type of simulation is not intended to predict the future. It is an exercise to see if the transfers last for the year will the poverty rate rise. The answer I find is no. Beyond that I can’t tell what will happen. BTW, the reform was “implemented successfully” in that energy consumption came down along with the poverty rate. The 2010 reform, defined as increase energy prices and offer cash as compensation is the most significant reform program of the last two decades. The fact that it did not last longer that couple of years means that the reform failed but does not make is a bad idea. Marriages that end in divorce are not necessarily bad ones from the start! Things happen.
2. This is a ceteris paribus exercise, not a prediction of what happens years from now. Yes, inflation can rise and many other things can go wrong, but they would not negate the fact that cash transfers are better the forex subsidy or that doing nothing is better. The government earns 200,000 rials per dollar for the food and medicine subsidy. If the payments are less than this, inflation should not increase by much on this account. As I note in the article, the calculation of earnings is much easier in this program than AN’s.
3. If they cut the transfers, and spend their savings on less useful projects, bad things happen, like rising poverty. I hope they keep the transfers.
4. People liked AN’s program. I have never heard that people resent replacing energy with cash transfers. And what would be the logic in that, especially for the poor?
Thanks much for taking the time and the helpful explanations. Perhaps it’s best to talk in person. Jut a couple of points for now.
First, I should have made it clear by success I mean ensuring that relative prices are broadly corrected and on a lasting basis. In that sense, I thought there was agreement that AN’s reforms were not successful (the causes being underestimation of the impact on inflation and fiscal, and limited supporting reforms). I don’t know about marriage, but I think the seeds of failure (in my sense) were probably sown when the sermon was being delivered! Your focus is not relative prices, it is poverty, which as you say for a year or so was lower.
Second, I’m not clear why you distinguish prediction and simulation here. But let’s not get into that. Either way the assumptions have to be self consistent. A simulation doesn’t mean one can input just any number for an important endogenous variable like inflation. To me as a casual observer 40 percent may be too low given the reforms.
So I feel more caution is needed in assessment if they haven’t estimated the impact on inflation and fiscal properly, the transfers are planned only for two months, and supporting reforms (ER, supply side, etc) are not carried out.
Not sure what was re-sent?
Djavad jan, what was the reason to re-send this blog at this time when irenewed efforts appear to suggest the initial program 10 years ago was not really successful? Thanks