My op-ed in Mint: Freak games mandarins play

The folly that is the exit tax

The argument against the exit tax was published in today’s Mint. (They changed the title.) If you have comments why not drop a note to the editor?

Two other posts that you must read on the topic: Gulzar’s post on Urbanomics, and Mukul Asher & Amarendu Nandy in DNA, over here.

27 thoughts on “My op-ed in Mint: Freak games mandarins play”

  1. Nitin,

    Don’t know if comments on the op-ed are welcome here. Didn’t feel like mailing anyone so am using the space here. I found several parts of the article quite jarring.

    1) “But in 2006, remittances through formal channels alone stood at $26 billion (equivalent to 3% of the GDP). Around two-thirds of this arrived from the US and Europe. In comparison, the Central government’s higher education budget for 2006-07 is just over $1.1 billion. Not at all a bad return on investment. “

    By that way of comparison, should we compare our total domestic economy with the total government expenditure on education, and feel happy about the ‘return on investment’ ? A significant amount of that $26 bilion comes not from IIT-IIM graduates but from Gujarati businessmen who have not recieved any benefits whatsoever from the subsidies on higher education in India. Will the return on investment be infinite in their case? Not everyone who will be charged the exit tax is contributing to the remittance, not everyone who is contributing to the remittance will be charged the exit tax. There’s no point comparing two such loosely dependent random figures and concluding anything from it. If GoI was to charge every Indian who left for foreign shores this exit tax, you would have had a point with those figures – but not in this case. In any case, the $1.1 bilion is borne directly by every tax-payer. The $26 billion is remitted to a small subset of that. Any benefits that accrue to this larger community of tax-payers are circuitous and unassured – ‘in the long run’ types.

    2)” So how does the government set the quantum of the exit tax? Given the externalities, merely recovering the subsidy (i.e, the difference between the fees charged and the real costs) underestimates the actual loss to society. How then does the government price the “benefits to society”? Well, it can employ economists and accountants to calculate these costs each year, adjust for inflation, exchange rates, financial status of the family and, yes, the graduate’s caste, too. ”

    Incorrect. Possible positive externalities are never counted as opportunity costs. The government does not need to count the benefits to the society – it just needs to recover the subsidy along with the interest to be at an economic no-gain no-loss state. Exit indemnities are not such unheard-of freak mandarin games as you think them to be. The government of Singapore sponsors the education of some Indian engineers every year through the Singapore Airlines scholarship. The scholarship (along with the tuition grant subsidy) is worth a total of about SGD 30,000 a year. The total indemnity amount, at the end of 4 years, is about SGD 140,000 – simply a calculation of principal + interest for each of the 4 years. This can be paid back by working for 6 years in Singapore (the scholarship carries a bond) or by leaving at some point of time in between and paying off the remainig indemnity. The method remains a simple time value of money concept, without thinking about the possible positive externalities. You have imposed your imagined scenario on the GoI – does it actually intend to do that?

    In any case, you seem to believe that the positive externaities are being enjoyed even without the tax ($26 bilion vs $ 1.1 billion etc). In such a scenario, why are you forcing the government to think about the externalities involved at all?

    3) ” Most importantly—and this is where the Israeli day care centres come in—the tax is likely to be counterproductive, and cause many more graduates to leave and never return. That’s because once the moral incentive to stay is converted into a financial incentive—at a small fraction of the graduate’s salary abroad—it can be paid off without any “guilt”. “

    Let’s transcribe your argument to another scenario. Once the moral incentive to join the civil services is converted into a financial incentive – at a small fraction of what such a capable student may recieve if he/she joins an IIM etc – it ceases to have any value. Hence, the government should not think of raising the salaries of civil servants. Makes sense to you?

    The flaw, of course, is your view that the incentive is getting ‘converted’. Do you think that a person who has a moral incentive to stay in India will suddenly start feeling bad that he is not being charged a tax that other people are and hence the value of his incentive is lesser and hence he should leave and never come back? Do you think that a person who wants to go abroad and eventualy return is disincentivised from doing so simply because the government recovered the costs on the subsidy that it offered him – which, as you yourself mention, is a fraction of his salary abroad ? If I can take a huge pay cut and either stay or come back, will the additional ‘fraction’ mean too much to me – when I probably myself realise that I did not have the claims to that additional fraction in the first place.

    4) ” In today’s economic conditions, there is no reason for the government to subsidize higher education, certainly not at IITs and IIMs. There are numerous means to finance the education of students who cannot afford the fees without having to resort to subsidies. There is no reason for it to sit atop an artificial shortage of doctors, nurses, engineers or management graduates. “

    This I genuinely cannot get. You suport no subsidies in higher education, and ideally that’s the way it should have happened right from the begining. In that case, there would have to be a large corpus – partially public and partially private – that would sponsor the higher education of students in need, mainly through loans. (And possibly some grants, the US model basically) Such a fund/corpus/institution would recover its costs by some repayment methods. I presume that you have no problems with such a recovery, and yet you have a problem when the government is trying to move from ‘no recovery’ to ‘partial recovery’? At best, you could have said that such a tax is discriminatory – the government should recover costs from every student and not just the ones who are going abroad. The grounds then should have been fairness and equality, and not incentives and disincentives.

    Even in this case, I would argue that the MBA who stays back in India and cuts (say) the New Delhi Power Ltd costs by Rs 10 crore is significantly more directly beneficial than the one who leaves for the US and then remits Rs 10 crore back home, which may or may not get invested in business here. The contribution made to the economy by being a part of it is simply not equivalent to an equal monetary contribution sent as money from abroad – positive externalities may explain why. So, the government probably has a strong case for making the staying at home/ going abroad distinction between those who use its subsidies.

    Fundamentally, I believe that your premise that the government is trying to incentivise people to stay in India or to disincentivise them from going abroad is wrong. It is simply trying to move from 0 to something in terms of the costs that it recovers – irrespective of how an incompetent HRD minister may present his case. And if an IITian’s demand for going abroad with respect of this recovery (a fraction of the dollar salary that he expects, remember) is rather inelastic (as I think it is), then it is an economically smart decision by the the GoI.

  2. Dear Ritwik,

    You should have mentioned your last point first…would have indicated what you were driving at. You say it’s an economically smart decision by the government, because it’s moving from 0 to “something in terms of the cost it recovers”. It’s financially smart in a sense, but it does not answer the question why the government is spending money in higher education in the first place. In fact, as I have written, it would be smarter to move from 0 to the entire $1.4 billion…by not spending it on subsidising higher education.

    On the ROI bit—it’s not my idea to compare ROI in higher education. I merely do so because that is the rationale used by the proponents of the exit tax. So if you think this rationale is incorrect, then you are shaking the very foundations of the argument for exit tax.

    The government does not need to count the benefits to the society – it just needs to recover the subsidy along with the interest to be at an economic no-gain no-loss state.

    And why is that? So the Singapore government has a particular way of doing it. That’s neither here nor there. Do you know their expenditure on primary schooling, their enrollment rates, completion rates etc? Without looking it up, we can safely say that it’s very very high compared to India. The priorities of their educational spending is different. Their constraints are different. In India’s case, where budget constraints are severe and basic education has fallen apart, the marginal rupee is best spent where marginal benefits (direct benefits + externalities) are higher.

    Your normative position (cost-recovery is all that is needed) cannot answer the allocative question: given cost-recovery is possible for several courses, where should India invest its education rupees? In training fashion designers or doctors?

    Do you think that a person who wants to go abroad and eventualy return is disincentivised from doing so simply because the government recovered the costs on the subsidy that it offered him – which, as you yourself mention, is a fraction of his salary abroad ?

    Yes. I think so. I’d say that the moral incentive to return is weakened or destroyed. That’s what Levitt and Dubner suggest in the fist chapter of Freakonomics. (Your other questions are either digressions or I can’t understand them).

    On point (4) you make too many presumptions. All I’m saying is that the state should stop regulating (managing supply) and subsidising (managing demand) of higher education. In this op-ed I make no arguments about how higher education ought to be funded.

    Indeed, if the government stops subsidising IIMs, it’ll save us the need to go into complicated (and wholly unnecessary) calculations of who contributes more (the MBA @New Delhi or the MBA who sends remittances) and hence deciding whether or where to recover the subsidy from.

    Finally—there’s a difference between financially smart and economically smart. A government that balances its budget or generates surpluses by cleverly levying taxes may be financially smart. But that does not mean that it is economically smart (Public Finance 101).

  3. Nitn – “In any case, unless the government can guarantee the graduates equivalent employment, it has no moral basis to stop people from seeking a better living abroad.”

    In that case all Indians who the Govt. has failed in one way or another might as well give up and try emigrating.

  4. Anuj,

    That does not follow from my argument. The subject of my argument is the state. It is states like East Germany and North Korea prevent their people from pursuing better opportunities abroad.

    But, in itself, I see nothing wrong with what you suggest. Not “might as well”, but “have the right to/would be justified in” (See this).

  5. Nitin,

    Don’t know if we will get anywhere with this, but let’s see.

    1) The decision is economically smart because I believe that the demand to go abroad is inelastic – hence you can raise prices etc. (in this case charge a tax – in my opinion the Rs 800,000 + interest over some amount of time that such a student will be required to pay in total this will not really deter an IITian from going abroad.) Thus, the government generates extra revenue. The remittances will still continue.

    2) “The justification for the tax is the recovery of the huge government subsidy for tertiary education.
    Once the moral incentive to stay is converted into a financial incentive, more graduates could leave India
    The presumption, of course, is that keeping the graduate in India is the only way of getting a return on the government’s subsidy. ”

    Do you notice how fluidly you move from ‘recovery’ to ‘return on investment’? Indeed, the rationale for exit tax (as I see it – I don’t really know the exact statemnets of the HRD ministry) is recovery, not return on investment. And even if we were comparing ROI in education – the $26 billion as foreign remittances vs $ 1.1 billion as investment in higher education are just not the right figures to use. Education is just one of the variables that affects total productivity – taken as the sole varaible, ROI will be something like 50 times because the national expenditure on education is around 2% of GDP. Those figures served no puprose, whether one believes in the RoI argument or not.

    3) I haven’t read Public Finance but I know that levying taxes to increase revenues is an economically unsound decision – that is simple microeconomics, really. But that is so because the taxes affect productivity in the economy. Here, the charge is being levied on someone who is leaving the country. Do you have a problem with the ‘tax’ bit? Would it be ok if the government simply said – ok, if you want to go abroad, pay up the subsidy we offered you in equated monthly instalments of rs X for Y months ? How can you call recovering your costs a ‘god financialy, but bad economically’ decision? Not recovering your costs is the worst posible economic decision. Any move from there is good.

    4) ‘Cost recovery is all that is needed’ is not my normative position – it is standard economic practice to not count possible positive externalities while calculating opportunity costs. This is so because the externalities involved are specific to the enterprise and are not assured if the invesment is made elsewhere and hence it doesn’t make sense to include them in your opportunity costs. What I said is that you presumed that the government will sit down with economists and accountants and calculate all the external benefits of education by some long process and spend large amounts of taxpayer’s money while doing so. This is a purely imaginary scenario because that’s never how it happens. I mentioned Singapore to substantiate my point because I know all the details involved there. The process and method of calculation remains the same everywhere – even in the Federal Financial Aid offered by the US government. At max, the rate charged may be the interest rate and not the rate of inflation (to ensure not just cost recovery, but some return on the investment as well) – calculation of positive externalities of education is just not done. You imagined a scenario, pointed out its flaws and criticized the government on the basis of that scenario – all this becomes irrelevant when that situation is not going to happen in the first place.

    5) Does Levitt’s theory on incentives hold true in my example of the salaries in civil services? An imposition of a fine is a punitive measure – that’s just not the same thing as a recovery of valid costs. I think that the moral incentive comes from a sense of patriotism, not from the fact that the government sponsored my education. For those to whom the moral incentive matters, it will stay regardless of the tax. For those to whom it doesn’t, we are indifferent.

    6) Yes, the money is better spent on primary education. Yes, we should have never subsidised higher education in the first place. But you’re not just saying that government should get out of higher education – you’re criticizing a specific initiative of the government. Given that we are in the business of subsidising higher education, is partial recovery of subsidy a worse scenario than no recovery of subsidy? You claim that it is – I fail to see why. And is any of the presumptions I make about the financing of higher education invalid? Wouldn’t we need a huge corpus? Wouldn’t this corpus need to recover the loans/subsidies that it offers? Isn’t it true that the institution that manages this corpus will impose a similar condition on its benficiaries – either work for me or pay up.

  6. Dear Ritwik,

    1. You are using economics terms (inelastic demand), but still are confusing financial ‘smartness’ for economic smartness. Raising revenues does not mean economic smartness.

    2. Recovery is a return on investment.

    3. Levying taxes is not always an economically unsound decision (think about negative externalities). The rest of that paragraph is facile. Calling a tax something else or splitting it into instalments doesn’t change the argument.

    4. I don’t know what you mean by ‘standard economic practice to not count possible positive externalities while calculating opportunity costs’. But if you are ignoring externalities, then you are very likely making a mistake. And if you do wish to support your case with examples, please use those relevant to India.

    5. I can’t comment on Levitt’s theory and civil service salaries.


    Given that we are in the business of subsidising higher education, is partial recovery of subsidy a worse scenario than no recovery of subsidy?

    It depends on the cost of recovering the subsidy. The marginal benefit of recovery must be higher than the marginal cost of collection, for the whole economy. Even taking a subset, the cost of collection must be lower than the collected amount. You might be surprised to know that in India, for income tax, for many years, the cost of collection has come close to, or exceeded, collected amount.

  7. Nitin,

    Looks like we’ll have to agree to disagree. Anyway,

    1. I don’t understand why the decision is economically unsound. Financial smartness may be different from economic smartness but the’re not mutually exclusive, right?

    2. I assumed RoI to mean positive returns. Anyhow, this is a question of semantics.

    3. I know about negative externalities. Ignored that case as it’s not relevant here. And I spoke of breaking it up into instalments because that brings it conceptually closer to a loan – which is what a subsidy essentially becomes if it is being recovered in the future.

    4. I am not ignoring the benefits of positive externalities – I’m only saying they are not counted or estimated as costs when a loan/subsidy is to be recovered, because they’re not costs in the first place (which is why they are externalities).

    5. And I think it’s inapplicable. To civil servant salaries as well as the proposed exit tax.

    6. Agreed. An assertion about the marginal cost of recovery vs marginal benefit of recovery would be premature either side of the debate, right?

    In any case, I think we’ll have to let this be. I have an end-term tomorrow and you must surely be busy. Thanks for engaging in the discussion.

  8. This is another hare brained, populist solution from the brilliant minds that seek to ‘govern’ us.

    Nitin, I think you ask the right question – why does the govt want to continue to subsidise elite institutions in the first place? If recovering the subsidy cost is the answer, why not eliminate/reduce it in the first place? Going by Ritwik’s instalment scheme argument, what is stopping the govt from making students pay the full cost and provide instalment options for those who cannot pay immediately? I don’t think it’s anyone’s case that a graduate from one of the premier institutions will be unable to land a decent job within the country should they choose to do so. In any case, how does a graduate working for IBM/Reliance/Goldman Sachs in Bombay or Bangalore ‘benefit the nation’ more?

    This is one of those policies that everyone seems to like until they are at the receiving end themselves. The fact is a whopping majority do not believe they benefit from it (and justifiably so), so it is easy to go after the golden goose. What people also forget is that higher education is subsidised from top to bottom. I passed out from a modest university where I paid less Rs.100 towards living expenses (this may have changed). Even after a substantial tuition raise, I’m sure we were still paying a fraction of the real cost (and for that I do thank the lord!). Graduates migrating overseas are an easy target with no way of escaping the system, just as the salaried classes are easy targets when it comes to widening the direct tax net. More patchwork on the quilt, that’s all this is.

  9. I also believe that the so-called ‘moral’ incentives are highly overrated. People often make more rational economic decisions than we are led to believe. More than any imagined love of country (not that there is none), it is really family and cultural belonging that seem to keep people in the country (among those that have a choice).

  10. Nanda Kishore,

    On comment #10, response to moral incentives form part of rational economic decisions. And yes, these include not just (or not even) ‘love of country’ but all the other things that you mention.

  11. Ritwik,

    Disagreement is fine: but you’ve missed the point. You must first ask yourself why should a government fund/subsidise anything? If it’s only to make financial returns (whether simply recover costs or maximise profits) then there are umpteen ways it can do so—like run lotteries or make equity investments etc. Why go through the hassle of setting up colleges, employing teachers, subsidising students and then collecting back the subsidy?

    The exit tax is economically unsound because, firstly, at best, it does not prevent people from leaving (which is the main problem). At the worst, it might encourage more people to leave. Second, given inefficiencies, wastage and corruption, the economic cost of recovering the subsidy is likely to be exceed the total amount of subsidies recovered. Thirdly, as you’ll see from the link to Asher & Nandy’s article in DNA, the entire problem is defined wrongly.

    Indeed, what you think is premature is central to the discussion on any tax (even if you think the exit tax is somehow a good idea). It’s not assertion. Here are some facts:
    Estimates of compliance costs are high by international standards even for salary earners and excessive for non-salaried taxpayers equalling, at the median, 130 percent of taxes paid. (Compliance Costs in developed countries varied between 3.91 percent of tax revenue in the UK in 1986-87 to between 7.9 to 10.8 percent in Australia in the 1990s.)

    The overall social cost of the personal income tax, adding administration costs and subtracting bribes is 60 percent of tax revenue. [Das-Gupta 2003]

  12. Nitin,

    I dont think people will be prevented from going abroad en masse because of this ‘tax’ – and you are right about that. But you are completely missing Ritwik’s point as I get it (and which in any case is my point) – precisely the fact that people will not stop emigrating en masse is what makes this policy quiye effective (all the standard inelasticity jargon)

    Basically the govt. is trying to recover (partial) costs from one subset of students – those who go abroad, and not from those who chose to stay home. So when Ritwik says that perhaps people have a problem with the tax terminology per se, he doesnt seem facile to me.

    So as I see it, if the govt. recovers costs from those who stay in India too – then your opposition will be lessened. You will continue to harp about the bogey of corruption exceeding recovery – but its just a bogey as you have no data to prove that.

    Hence what is EFFECTIVELY happening over the lifetime of an IITian etc – is that he is getting a scholarship to stay home. Now we dont know whether that will help India’s economy vis-a-vis more remittances if there was no such tax (and I am agnostic about this point unlike Ritwik).

    BUT since you have already conceeded that the flow wont stop, that point IS MOOT.

    So if you cant prove the “corruption/inefficiency exceeding recovery” bogey, and if you agree that the human capital flow will be largely the same – everything else remains more or less THE SAME (including future remittances) except that the government gets more revenue.

    Basically you are taking an ideologically libertine position rather than a pragmatic one – and you have my sympathies for that (just that though). I agree that given the lucrative professional careers awaiting IITians, IIMians etc – there should not be less/no subsidy in the first place. But since thats not coming, this solution is the second best one – and definitely better than no step.

    Positive externalities are there of the IIT/IIM education, NOT the IIT/IIM subsidized education – given the present day labor market (in)elasticities. I would support an exit tax for all students – not just outgoing ones, and I would support then to dismantle the systmem and officially have student fees – and hence loans – increased.

    But till the subsidy is directly withdrawn, this is a good step. Your article is quite simply faulty

  13. I meant in my previous remark that “I would supprt an exit tax for all current students – not just emigrating ones (instead of “outgoing ones”)”

    In fact, if legally possible – even retrospectivey. Though that wont come true.

    You see this policy as trying to prevent emigration. But Nitin – this policy is just recovering costs (without lessening the positive externalities)

    And I see, you have posted that maybe even 60% is the cost of tax-collection. Now thats what I think is facile – no time to get into details now. But even if its true – it is still less than 100% and my point remains valid.

    My only exogenous variable was that emigration would be inelastic – but I repeat that I dont need to prove that to you at least, as you yourself have conceded it.

  14. “Once the moral incentive to stay is converted into a financial incentive, more graduates could leave India ”

    That is quite frankly ridiculous – along with the Freakonomics analogy. You suggest that now since people can “buy” working abroad – their guilt will decrease and they will be more inclined to do so. Quite a leap of imagination – and I am happy that no one in the HRD ministry has heard of Uri Gneezy and Aldo Rustichini. In that Lewitt case, parents in a way know that they are paying for their children’s security for extra time. Its not about guilt. Earlier they could not have held it against the school authorities if they were late and something had happened to their child. But now they believe they have a better chance at that. Because – after all they are paying – call it a fine or tax or simply overtime charges.

    In this case (and why dont you give an Indian analogy, the way you ask Ritwik), IITians are not being fined. They are just not being given a scholarship/subsidy. The cases are very different. Any elasticity (because all that moral stuff seems gibberish) will be somewhat towards staying home – not catching the next flight to the States just because “Hooray, we have bought off our guilt”. In the Indian society there is no ostracizing emigres – rather they are sought after in all respects.

    Plus you seem to suggest that remittances help India – and I agree with you. As I said earlier I am agnostic about the “what-ifs” regarding had they stayed back.

    But you implicitly assume that the IITian doesnt know about his contribution to India by going abroad – remittances / Silicon Valley or Wall St. pride about Indians etc and even resident Indians in general being more respected because of his job. Their might be personal guilt – family based and even patriotic. But how many IIT/IIM people will have that kind of guilt lessened just coz they paid some money to the government. I have many friends/relatives in IITs/IIMs and I myself got into IIT – though chose not to attend. The answer is not many.

  15. Dear Harsh,

    You’ve merely repeated Ritwik’s arguments. Please read my responses to Ritwik, especially comment #12. Read the comment, and follow the given links, before you declare the argument is a bogey and there is no data to prove that.

    Surely, no one disputes that it is a good idea to levy taxes where demand is inelastic. But that does not mean you levy taxes on everything that is inelastic, even if it brings in revenues. Extortionary states do that—like the British taxed salt. Besides, if revenue is what you are after, why not say increase taxes on cigarettes, private jets, polluting vehicles and factories, and make the same money. (It’s more efficient, because increase in compliance costs will be marginal). Or get politicians to move out of state-owned prime real-estate that they occupy in middle of big cities, and lease it off by auction. Why tax graduates? You can’t call it “clever” or “pragmatic” just because it raises money for the government.

    As for your opinion on Levitt, Gneezy and Rustichini, it’s just that—your opinion. Besides “friends & relatives” arguments may make good conversation, but are generally not a basis for public policy.


    Please see Das-Gupta’s paper. Compliance costs are the costs to the entire economy (costs incurred by the taxpayers, costs incurred by the tax collectors, wastages, minus corruption). The government figures just tell you how much money the government spent on tax collection. These are two different things.

  16. It does not raise money for the government
    It recovers money. There is a difference.

    And yes, I very much intended to repeat Ritwik’s argument – just in case you would understand. And my friends/relatives’ anecdotal evidence is over and above my argument.

    But thats ok, perhaps I am wrong – debates are meant to enlighten people and perhaps change people’s opinions, not further entrnech egos. If only more people understood it.

    Anyways, a last dash: Why recover the true cost from graduates (something you agree with but strangely dont like it if its post-college…)? To send out correct signals. Lets see…

    Our academic system should produce people with high creativity and adaptability. Technincal/management specific skills are needed but so are pure sciences/social sciences graduates in larger numbers (I wont even make a case for humanities). When the benefit (salary for engineer/manager etc) and cost difference in a lopsided subsidized system exists – more people want to be engineers/managers.

    Its heartening you agree on at least that – let us stop subsidies, but for current students/recent graduates – why not recover the costs. Why let more people into the IIM etc mania than is required. In India, MBAs are entry level professionals mostly rather than midcareer guys/gals – as it should be, because then the bachelors education would have to be more accountable. In India, where people of your ilk look for a Free-Markets Mahatma (and I fully fully support it) a more liberal arts and sciences education is needed rather than a vocational one.

    Forget revenue – you say. Aye Aye sir. But what about the education system. Isnt this tax simply a negative subsidy. I havent perused ur arguments/articles but I have glanced through them – but I still dont get what your hangups about this policy are.

    I agree there are some inefficiencies in this system – and I am the first guy to support simply ending the subsidies for these elite institutes (except for R&D etc. obviously) – and then IITs/IIMs will become truly world class on the faculty/infrastructure front too and not just known (justifiably) for their brilliant students.

    Basic economics. Just to send out the write signals to the education system too. Coz even with this tax no positive externalities are being reduced

  17. Nitin,

    1. Recovery of a subsidy is not extortionary – especially not if one believes that the subsidy should not have been offered in the first place. I don’t know why you brought in the example of the salt tax. The inelasticity of demand simply means that economically, the decision will not backfire – IITians will continue to go abroad. They will still have families and financial interests in India. They will continue to send money back home.

    2. I also don’t get your “either this or that” position wrt tax. It is true for higher education vs primary education (yes, given a certain number of education rupees, we’re better of spending it on schools than on IITs, because it is expenditure of limited money – an either this or that case)but it is not true for exit tax vs tax on cigarettes. One can make the ‘same money’ by increasing tax on cigarettes, and make some more money (or rather, recover some money that one has already spent) by having an exit tax – the extra effort and costs are justified if the money recovered are more than the costs. X + Y – (Y-delta) = X + delta > X. This ‘making money’, as I crudely put it, is a ‘financially sound, economically unsound’ decision only if it is likely to effect negative changes in the emigration / remittance/ moral incentive pattern which will wash off the benefits of the delta.

    3. I disagree with your argument on moral incentives. Neither of us will be able to prove our stands on this, so let’s leave it.

    About emigration, you say that at best it doesn’t stop people from leaving – nothing will until the economic opportunities become comparable. As a democracy, we also do not wish to stop people from leaving. All we are saying is – if you leave, give us back the money we spent on you. You also claim that at worst, more people will leave. I disagree with your stand on moral incentives and so don’t think this will ever be the case.

    Remittances are made for personal and business reasons. As long as there are families and financial interests back home, the exit tax will have no effect on them.

    4. Thus, the only point that remains is – what exactly wil be the costs of collecting such a tax.

    Let’s see what I dug up from the same article that you link to. (excessively long analysis alert)

    ” Information base, sample size and response rate: Besides primary survey data and case studies
    relevant secondary data are used. For the survey, from a list of 10,234 names the eventual response
    rate was a disappointing 2.36% amounting to less than one taxpayer per 100,000 income tax payers in
    2000-01. Overall, the final sample of 172 taxpayers is biased towards high income respondents and
    salary earners. Therefore, findings of this study must be taken as very preliminary and subject to a
    large error margin. Nevertheless, the high compliance costs documented make it clear that a reliable
    cost assessment is urgently needed. ” – ridiculously low sample space

    “Universe Coverage: According to the report on direct taxes of the CAG (2001), the number of income
    tax assessees stood at 1,95,67,937 in 1999-2000. This implies a sample coverage of 0.00089 percent
    or just under one taxpayer per lakh (1,00,000) of assessees. Overall, the sample is biased towards high income respondents, salary earners and possibly, the highly educated.” – sample bias.

    Let’s take a look at some data.
    (from table 8) : average monetary compliance costs : 58.7 % of taxes paid, 1.67 % of income
    : median monetary complance costs : 14.1% of taxes paid, 0.26% of income

    obsevation 1 : huge difference betwem median and average. Not only is the sample biased, it containes tremendous outliers within itself.

    observation 2: 58.7 % of taxes = 1.67 % of income => average income tax actually paid = 2.8% of average income. In a country with progressive income tax, tax slabs of 10%, 20% and 30%, and a sample that is skewed towards the affluent, the tax paid is still 2.8% of the income. This means that through employing chartered accountants and tax planners, the respondents have availed enormous tax benefits. If the payments made to CAs are counted as compliance costs (because we’re measuring costs to the society and not the government), shouldn’t these benefits be considered ‘compliance benefits’ and deducted from the analysis?

    What’s more, the paper itself mentions that the average income tax paid is 8.21% of income (table 2)- widely different from our calculations. There are several such mismatches and the paper itself mentions a caveat that explains why – apparently, many people (even out of the 172 who filled the survey) did not fill the survey totally – many items were missing for quite a few respondents.

    Most significantly, 13% of the respondents were under legal tax scrutiny. This represents the actual bias of the paper – legal compliance costs form the largest share of the calculated compliance costs and in a more balanced representative sample, the results may be ‘wildly different’.

    However, presenting a detailed critique of the paper is not my aim, much as I am tempted to do so. Let’s just say that the paper is tremendously insufficient for statistical inference. Let’s also agree with the basic qualitative understanding of the author – that compliance costs are much higher in India than in developed nations and the tax system needs an overhaul. The author claims compliance costs of 60% of tax collected, and infers that at the very least even after major variations, the figure is likely to be much higher than the 3-10% of developed nations. You, on the other hand, extend this 60% to almost 100%. How fair is that?

    Lastly, income tax laws are convoluted and hence involve tremendous compliance costs – in an exit indemnity agreement based on the time value of subsidy being offered (a more or less fixed amount, subject to not to many clauses and sub-clauses, more like a loan agreement than income tax), complaince costs are not likely to be remotely as high.

    In the light of all this, how correct is your asertion that the costs of recovery do not justify an exit tax?

  18. Dear Harsh,

    If you follow my argument (in the last few paras of the op-ed), I argue that the government should deregulate and get out of the higher education sector. It has no business sending any signals “right” or otherwise…that too by taxing graduates. For someone who professes to be in favour of free markets, I’m troubled when you say

    “Why let more people into the IIM etc mania than is required.”

    It’s neither my call, nor yours nor the government of India’s to determine the ‘required’ number of people.

    And it’s interesting that you mention that you favour charging the exit tax on current/recent graduates. Why not every graduate since the system of subsidies was introduced? There’s far too much arbitrariness here, don’t you think.

    (Btw, you don’t need to repeat arguments made by someone else, especially voluminous ones. I read them)


    Let’s say, only for the sake of argument, that cost of recovery is not an issue. Your only argument is that an exit tax is good because it helps recovers money. But then I already gave you a way to recover even more money: don’t spend it at all.

    Let me invite those interested to offer closing arguments—no new points, just summarise (in as few words as possible), why you support/oppose the exit tax.

  19. Ritwik,

    Shorn of technical jargon, your argument is: Government should because government can!

    At the risk of repeating what Nitin has already said, the point is fairly simple,

    a) The government across the board subsidy of higher education is wrong, attempting to recover a certain% doesn’t make it right. In fact, it creates a perverse incentive to continue down this path because–well, we are are recovering some money. In fact, I am very curious to know: How much money government hopes to generate from merely taxing IIT/IIM graduates going abroad? Yes, it is proper to use tuition refunds to redirect people to areas where they may be needed–rural communities for an example. But making a distinction between people who choose to move abroad and those who stay back is stupid.

    b) There are immense technical difficulties involved. For example, if someone goes abroad for a few years, does he still pay graduate tax? How many years? What status? In a country where a vast majority of people don’t pay any taxes, it is inconceivable that government would be able to fine tune exit tax to a level where the cost is lower than recovered amount.

    Finally, the concept of graduate tax it self is not without dispute, See this report by Lord Dearing

    A graduate tax is attractive because it has the potential to secure large additional resources for higher education, but it provides no means by which individuals can pay their contribution upfront, and thus does not deliver additional funding in the short term. For the graduate, it is open-ended, resulting in those who are particularly successful being expected to contribute large sums. For institutions, it would not guarantee that the income from the tax would benefit them because, to do so, would cut across the general principle that tax revenue is not earmarked for particular services

  20. Since you’ve asked for a short summary – I support the tax because

    1) I don’t think it’s a tax at all – it’s more like a loan recovery.

    2) We’re moving from highly negative to somewhat less negative. In the absence of any possibility that a move from negative (subsidizing higher education) to positive (more subsidies for primary education) will take place anytime soon, I think this shift is good.

    Apologies for the terribly voluminous previous comments.

  21. Deregulation is badly wanted. Nitin is right on that. I was just saying that since the HRD morons wont do that, this is OK for now.

    Obviously since Nitin’s position is the correct one (not the ground-reality one): he can ask me the “what-ifs” and the “why-nots”

    But I will repeat the last time – Nitin is right. But so is Ritwik – though definitely less so. And since Nitin’s advice hasnt been followed yet, let us not oppose the Ritwiks etc.

    Is there a “the end” emoticon, guys 😉

  22. Clarification to Nitin:

    when I said “numbers required” regarding IIM type guys, I meant as determined by the labor markets, not by me obviously.

    You are distinguishing between stay-homes and emigres. I am distinguishing between total subsidy and less subsidy – the latter not just the revenue, but also for better signals.

Comments are closed.