11th January 2014
A comprehensive demolition of the dangerous Arthakranti/BJP bank transaction tax proposal
[ADDENDUM dated 20 November 2016: Really bad days in store for India? Chidambram claims that Modi is implementing Arthakranti. While I don’t see the substantive part of Arthakranti promoters imposed yet, Modi would be a chump if listens to them.]
I have been in touch with the Arthakranti team over email and in other ways. FTI member Supratim Basu attended their Hyderabad presentation in late 2012. The Arthakranti team participated in the Haridwar National Reform Summit in April 2013 and presented a key outline of their proposal. While Surpatim and I disagreed with its claims (in 2012/early 2013), I agreed to make brief mention of it in the draft SKCF reform agenda with a commitment to get it reviewed in the future.
Since then I have been busy with other activities and more or less ignored the Arthakranti idea, given its obvious limitations.
However, the group is extremely enterprising (like AAP). After first having persuaded Swami Ramdev (a simple man with little understanding of economic principles), they’ve now gone and persuaded BJP – which is a bit of a shock! – given BJP had access to people like Arun Shourie, and should have known better.
Given BJP’s (current) advocacy [at least active consideration] of this proposal, I have no choice but to take time off to thoroughly demolish this proposal, once and for all. This blog post brings to bear almost all relevant arguments to rebut the bank credit (transaction) tax proposal.
Summary of the Arthakranti proposal
Before reviewing it, one should know the proposal. So here are the details:
- Withdrawal of existing Taxation System completely (except customs i.e. import duties).
- Every Transaction routed through a bank will attract certain deduction in appropriate percentage as Transaction Tax i.e. Single point tax deducted at source. (say 2 %). This deduction is to be effected on receiving/credit accounts only.
- This deducted amount will be credited to different Government levels like Central, State and Local (say 0.7%, 0.6%, 0.35% respectively).
- Transacting Bank will also have its share in this amount as the bank has a key role to perform (say 0.35%).
- Withdrawal of High denomination currency (say above Rs. 50).
- Cash transactions will not attract any transaction tax.
- Government should make legal provisions to restrict cash transactions up to a certain limit (say Rs. 2000).
Or, as summarised in this article:
Essentially, everybody would get credit and debit cards, and transactions would be taxed at 2%. The receiving party would be taxed and the revenue would be sent to the government with the banking intermediary keeping about 17.5% of the collected tax given that it has a “key role to perform” in the transaction. To eliminate black money, high denomination notes (say Rs 100 and above) would be eliminated. Cash transactions will not attract any tax but would be illegal beyond a certain level, say Rs 2,000 according to Arthakranti’s website.
I had provided a link to the original Arthakranti documents in my previous posts but these links have disappeared. To make sure that the links are permanent, I’ve uploaded the following two documents on my server:
Arthakranti is, at its heart, a Bank Credit Tax (BCT). Some call it the Bank Transaction Tax (BTT). It is intended to be a tax on the real economy, not on financial transactions. (Any Financial Transaction Tax of 2 per cent on stock/securities/forex trade would drive ALL capital out of India. However, there is a risk that BCT could apply to some financial transactions, thereby significantly dampening financial arbitrage and use of financial instruments for personal financial planning] .
WHY ARTHAKRANTI IS A BAD IDEA
1. Not based on ANY first principles
The Arthkranti proposal has no theory behind it that links its prescriptions to the social contract. Any tax model should start by asking: why should there be any tax, which types of taxes are appropriate, which are most practical, etc.
There is a huge discipline of public finance to guide us. Let’s use it. There is also a huge literature on public (collective) choice which throws useful light on this question.
I’ve outlined a theory of taxation in two articles published in Freedom First. I’d suggest you consider reading these first: A liberal perspective on taxes – Part I and A liberal perspective on taxes – Part II.
Any Tax ‘solutions’ not grounded in deep-rooted theory and empirical analysis are likely to make things worse.
The theoretically ideal is for a citizen to pay the full cost for all government services he receives. Where this cost can be easily calculated (e.g. user pays model) and collected at a low cost, a cost-recovery/ market-based recovery model is appropriate. An example is the charge for parking a car on a government road. That should not be raised through a general tax. But where attribution is either not feasible or transaction costs large, alternative methods of taxation are needed – including broad based levies and taxes. In general, the idea that there should be a single tax is pure nonsense.
When it comes to the broad-based (i.e. not user-specific) forms of taxation, the key theoretical construct is the ability to pay. The person most able to pay should pay more. Further, only individuals should pay, being voters. In that sense, a combination of wealth and income tax should be all that is needed. This generally involves a slightly progressive model, with most taxes being flat (i.e. proportionate).
However, a real-life taxation model is necessarily more complex, given the impracticability of identifying and taking into account all relevant income and wealth. Further, the ease of taxation by certain methods drives the use of these methods. As a result we end up with a mix of taxes (including user charges or fees and corporate taxes). No single tax can do justice to the complexity of human activity and interaction.
Revenue efficiency or the ability of a tax to maximise revenue collected at minimal cost is also important in the design of a general tax. Objectives that determine the ultimate tax mix of a country include:
– minimising the cost of collection and compliance;
– relatively low losses in consumer welfare per dollar of revenue raised;
– minimum deadweight loss;
– minimizing the (negative) impact on household decisions about where to shop, where to live, how hard to work, whether to save and invest;
– minimising the adverse effects of a tax on the incentives to undertake commercial and economic transactions;
– encouraging (or discouraging) business location in the country or state or county, or encouraging investment and job creation in general; and
– minimising annoyance to taxpayers (Jean Baptiste Colbert observed that “The art of taxation consists in so plucking the goose as to get the most feathers with the least hissing”).
(There are significant trade-offs between efficiency and other objectives: The most efficient tax is a fixed amount per person, a poll tax, but that is both impractical and severely iniquitous. Among the more efficient taxes is a broadly based annual land tax. It can also be combined with progressivity through a threshold and different rates based on the value per square metre of land. Transfer taxes on property (such as stamp duty) are inefficient and should be replaced with such a land tax which creates a minimal distortion on incentives to trade land.)
A bank transaction tax is attractive because of its ease of collection, but it fails on almost all other grounds.
The closest theoretical foundation for this tax comes from James Tobin who ‘proposed a global tax on foreign-exchange transactions in 1972’ [Source]. But note that Tobin’s tax was a FTT, where proportions are very small (the world has extensive experience of FTTs). It was NOT a transaction tax on the real economy.
Supratim Basu of FTI notes:
I am unable to understand the basis for this taxation – why 2% of bank transactions? Why not say, 5% of all rice sales in India? Or 3% of all wedding expenses incurred?, etc – there is no rational, logical basis for creating a tax structure in this form or manner.
Their starting premise is faulty – they talk about govt expenses as developmental expense and non-developmental expenses (putting defence and interest payments here), instead of talking about essential and non-essential expenses, as you would if you derive down from first principles of liberty and functions of the state. Then, they latch on to this whole concept of taxing bank transactions, because they CAN, it is easy and SIMPLE to track – without figuring out or explaining why such transactions should be taxed – and what is the theory behind taxing such transactions (which could be single or multiple or part of a chain where one leg is in India, the others are out of India, profitable or loss, as opposed to taxing income – they have some unheard of economists (probably cranks) who they are quoting (I assume not out of context) to support a banking transaction tax as somehow part of the “social contract”.
I am unable to see why we should not say, instead have a tax solely on bananas as a replacement tax for all the current taxes in the country?
These folks … do not realise that financial transactions for goods is actually a very small portion of the total amount of “bank transactions” conducted daily. Off-hand, I won’t have the numbers, but money transfers for the “real economy” are less than 10% of total transactions. On the balance 90% transactions the spreads are way below 2% – those transactions would simply come to a grinding halt, serving to send money velocity to zero, and our economy back to the 1700’s and village markets.
A tax in which the poorer (or less wealthy) sections of society pay the same share as the wealthy, is iniquitous (the idea here is simple: taking away 25 per cent of a poor person’s income as tax harms him far more than taking away a similar proportion of a rich man’s income).
The bank transaction tax is, at best a flat tax – thereby being regressive. [Note that a Financial Transactions Tax is broadly progressive because it affects only those who undertake such transactions, being generally the richest members of a society; to that extent a very small FTT can be potentially justified, but not BTT].
Supratim Basu of FTI notes:
A BTT has no connection with either the asset value or the income level of the person being charged – being a flat tax. By definition, this tax would have a higher impact on the poorer and middle class households, rather than the richer households, who would be able to pay for planning more efficiently.
3. Impossible to implement
Now, this is a fundamental issue. Arthakranti assumes that some legal mechanism can be found to force people to use cheques for amounts greater than Rs.2000. “Government should make legal provisions to restrict cash transactions up to a certain limit (say Rs. 2000)”.
As anyone with the remotest level of common sense knows, such laws are not worth the paper they are written on. The essence of public choice and economics is about understanding human incentives. Such simplistic solutions can never work. Elimination of the black market requires extremely fine understanding of the incentives of all players in a market. An example of how such false declarations can be removed from the real estate market is provided here. Without such reforms, the entire idea that people will use the banking system will fail.
People are always smarter than governments. They will find a way to evade the banking system. With bank transactions becoming a ‘nuisance’, the market will find a way to evade bank deposits entirely. Those with significant assets will now be able to hide easily.
Some of the things that people will do include:
a) Barter: One of the most likely and common ways of evasion is ‘netting out’ transactions – or, in simple language, barter. As the 1996 Inquiry into the Australian Financial System noted,
“In order to minimise FID liability, many corporations in Australia aggregate and net payments. This erodes the FID revenue base. Where it is not possible to aggregate or net payments, or firms are too small to justify the development of corporate structures to take advantage of the FID cap, FID could delay (or increase the costs of) the widespread acceptance and adoption of electronic payments in some markets.”
b) Anonymous transactions: Expect internet-based cryptocurrencies – where anonymity is the key ‘advantage’ to proliferate with Arthakranti. Hawala type transactions, backed perhaps by gold, will become common.
c) Use of gold and non-rupee mechanisms of exchange: Not just gold, but the use of foreign currency (e.g. USD/AUD, etc.) will increase. Smuggling of such types of mediums of exchange will proliferate.
In all cases it will be the wealthiest and smartest who will find a way to evade the banking system. No amount of Janlokpal will be able to unearth the increased black money.
Supratim Basu of FTI notes:
The reverse of the intended will happen due to such a poorly designed structure – increase in barter transactions, cash transactions, non-cash, non-traditional transactions (a buyer offered to pay me in gold for 30% of an apartment that I was selling in Pune, to reduce the stamp duty, and when I said I would not take a single rupee of cash!!)
They have to tell us how will everyone in India get a debit card when the formal banking system covers less than 45% of India’s population.
And as a commentator notes on this blog post:
It is highly difficult, almost impossible to implement in India for its predominantly cash trans. based economy and above all we have a dragon in parallel economy, which is double than our economy.
4. Anti-business, anti-economy
Transactions are the lifeblood of an economy. Arbitrage (trade) “oils” the economy and will be most severely affected. Speculators and arbitrageurs perform a critical function. They buy and sell in bulk across a wide range of goods and services (I’m assuming that financial securities and currencies are excluded from BCT except through net transactions). In this process, the most efficient use of any particular commodity is found and incentives set in place for the production of optimal quantities of goods and services. The most severely affected will be wholesale traders whose margins are wafer thin and who can’t possibly pay 2 per cent bank tax and still survive.
A BCT will create significant disincentives for arbitrage, trade and transactions (commerce), bringing the economy to its knees.
Supratim Basu of FTI notes:
A Tobian Tax, which this is both implicity and explicitly, has the effect of slowing down velocity of money/trades – the net effect would be to cause a slow down in the economy, possible sending it into a death spiral as people become reluctant to transact. Tobian taxes have been used successfully in so-called speculative markets, in the past – slowing down forex currency trading in France for example or our own securities transaction tax, which led to over 90% of the speculative trading to move from underlying to derivatives. So, there is historical evidence to show what Tobian taxes do.
Given that this is a tax on gross transactional value, the net effect on profits or income is likely to be of a magnitude higher than current taxation rates. The net effect would be to move these trades either offshore or to alternate markets, including barter, not covered by the BTT.
While investments that involve relatively few bank transactions or have a greater ability to access barter/ non-banking channels will be attracted to India, most other investments will flee India. Businesses have already been driven out of France due to the FTT, and out of Australia by the bank deposit tax.
The securities transaction tax is a little better – and indeed already exists in some form or shape, but Financial Transactions Tax which is broader is quite problematic: “The Chancellor has vowed to block any attempt to impose the controversial tax which has been blamed for driving business out of countries where it has been tried, including France.” [Source]
In Australia in relation to the Bank Account Debits (BAD), the following has been said [BAD is a tax on debits made to all accounts which have cheque access. Originally a federal tax, it was handed over to the States in 1990. It applies in all States and Territories except the ACT. This tax was abolished in mid-2010s]:
As the market for financial services becomes better integrated and international in scope, taxing financial transactions will become increasingly anti-competitive and inefficient. Australia is the only nation in the world to tax financial transactions directly. Its financial transactions taxes will not be appropriate in an environment where transaction costs and returns on investment are overriding factors directing the flow of investment and capital. Transactional business will be driven offshore to low cost transaction centres to avoid FID and BAD tax. Institutions able to offer low cost transaction services will enjoy a competitive advantage over their Australian-based counterparts for both Australian and international business.
Thus, 65% of businesses with annual turnovers in excess of $750 million already maintain offshore foreign currency accounts. 43% of this group cite FID as a major or decisive factor in this decision. One Australian corporate with large US denominated export revenue, established its foreign currency account in Queensland, as a result of which it saves $100,000 per annum.
Transaction taxes induce businesses to rearrange their affairs artificially and inefficiently. Thus 75% of businesses have implemented group structures specifically to minimise FID and BAD tax (rather than because these were the most efficient structures).
80% of banks consider that FID and BAD tax impede the development of new banking products. Thus, Westpac would like to introduce a cash management service which would automatically transfer funds from accounts in credit to accounts in overdraft. However, transaction taxes on the transfers would eliminate the saving.
65% of businesses occasionally or frequently change their treasury investment decisions because of the cost of FID, with the figure rising to 75% for businesses with annual turnover exceeding $1.15 billion.
FID is an explicit tax on saving. It is payable every time a customer makes a bank deposit. It also discourages savers actively managing their investments so as to maximise the return they receive, as every time a deposit is switched between accounts or between institutions, FID is payable.
The cost associated with transactions taxes reduces competition. For example, FID may make it unprofitable for funds invested with one bank to be transferred to another which offers a higher interest rate. [Source]
A study of the FTT (which is, as we have noted, far less intrusive than the BCT):
A Bank of Canada analysis of the effect of previous FTTs found that they tend to harm market quality, by increasing volatility, reducing volumes and raising the cost of capital. [Source]
Reduction in growth is an acknowledged consequence of the FTT: “Even the commission says that the tax will have a small negative effect on long-run growth.” [Source]
As a sensible commentator writes on this blog post:
Putting a tax on trade will depress the entire economy. In the present information age, the trend is towards a friction-less economy, i.e. increased trading, with very low overhead cost for each transaction. Putting a tax on transactions is like putting sand in the engine oil of a car. There was one comment that traders will, in this situation, increase their margins, and therefore would not suffer. It is true that individual traders can increase their margins to the extent allowed by the competitive environment. But the trading sector as a whole will suffer.
5. Will increase prices
With reduced trade/transactions and poor transmission of price data across the country, expect significant increase in prices. The lowest price or the most efficient form of production will no longer be feasible, as key price signals are switched off.
6. Practicality: Will fail to raise enough revenues
For this bit let me cite from this article which has researched this issue thoroughly:
At the meeting of the BJP’s brain trust, participants were informed that banking transactions are “merely a fraction of total financial transactions in the country with about 80% being in cash.” According to Mukesh Sharma of the Delhi chapter of Arthakranti, “if we account majority of these cash transactions into banking, this will amount to about Rs 40 lakh crore of annual revenue from this tax.”
Research by OECD economists Jorge Baca-Campodónico and Luiz de Mello and IMF economist Andrei Kirilenko, as well as a research paper for the World Bank by Patrick Honohan and Sean Yoder have debunked the claim that BTTs deliver the promised revenues. In a study of six Latin American countries (Argentina, Brazil, Colombia, Ecuador, Peru and Venezuela), the OECD/IMF team concluded that “for a given tax rate, revenue declines over time. Therefore, in order to meet a fixed revenue target in real terms, the tax rate needs to be raised repeatedly. However, we also find that successive increases in the tax rate erode the tax base by more than they raise revenue, and that the higher the increase in the tax rate, the more and faster the tax base is eroded. We conclude that bank transaction taxes do not provide a reliable source of revenue, especially over the medium term”.
“[T]he Latin American countries imposed these BTTs in addition to existing taxes. So, how much did they collect? Not much, is the short answer. In Argentina (2004), a BTT rate of 1.2% yielded revenues of 1.72% of the GDP. In Peru (1990), a 1.42% BTT rate yielded revenues of 0.89% of the GDP. In Ecuador (1999), a BTT rate of 2.0% yielded revenues of 2.51% of the GDP. By way of comparison, the dreamers in the BJP and Arthakranti feel we can generate 40 lakh crore by imposing a 2% tax on bank transactions. In case you are wondering, that is 40% of the GDP!
Need to keep increasing the tax rate
Once again, let me cite this article to make clear what happens when such a bad idea is introduced:
A study by Kirilenko in 2004 indicated that there was disintermediation (aka reduction in tax base) if BTTs are levied. And, remember, the longer you keep these taxes in place, the higher the rates have to be to generate a certain tax revenue. That means the tax base gets worse over time. So, we would have double digit fiscal deficits and an increase in black money within no time if the BJP’s proposal were to ever see the light of day. No wonder these taxes were abandoned time after time in country after country.
[Addendum, 17 June 2015: Also chanced upon this paper. “Arthakranti Plan: Noble Intentions but Muddled Thinking” by Parag Waknis, 2008 – haven’t had time to read it, but have transferred a copy on my server, in case the original is misplaced.]
The Arthakranti proposal is really bad.
- It will fail to raise the revenues needed
- Its incidence will fall on the poorer sections the economy (the rich will pay much less than they do today)
- It will reduce savings in the economy
- It will reduce investment and harm business
- It will increase prices
- It will create a massive black money/underground economy using hawala/ gold/ cryptocurrency
- It will harm India’s banking system by preventing it from providing a range of innovative products
- It will bring Indian economy to its knees
NO NATION IN THE WORLD has relied solely on a bank credit tax, and for very good reason. Countries which did use bank/transaction taxes for expediency have mostly abandoned/replaced them with more efficient and progressive taxes.
Worse, it is not a sensible way to deal with corruption
As Supratim Basu of FTI notes:
Demonetisation of higher currency notes has never curbed black money, including the two attempts in India.
Arthakranti won’t remove corruption, including from the real estate market. To remove corruption/black money, a range of other proven solutions (founded in the study of human incentives) are needed.
Chanakya was an expert on corruption. Let’s learn from him. BJP claims to represent the Indian tradition. Have its economists ever bothered to study Arthashastra – one of the greatest books ever written on economics and governance?
Overall, I’d give Arthakranti a “D” as a policy proposal. It says a lot about the policy skills of our major parties that BJP has adopted this dangerously ill-informed proposal.
What should the alternative system be?
Well, it should be based on first principles, and I’ve explained at length in the two articles in Freedom First, cited above. That does require a lot more knowledge/study than BJP seems capable of.
“But the country may get as finance minister Baba Ram Dev, who thinks all taxes can be replaced by a Tobin tax on cheque clearances. Little does he realise that the tax would have to be 25 per cent. If he imposes such a tax, everybody, including his acolytes, would take their money out of banks and settle all transactions in cash. But the economy would collapse much before that. After that, history would certainly look back nostalgically on Manmohan Singh.”
2) Sharad Bailur’s analysis in 2001
Arthakranti by Anil Bokil: An assessment By Sharad Bailur
Over the last several decades we have had a number of attempts to consider radical measures to revamp the tax structure with the aim of containing black (unaccounted) money and to give taxation a more equitable face, if not form. All have failed or not lived up to their promise up to now.
1. Abolish all personal taxation. Make Excise and indeed all indirect taxes progressive.
2. Have a system to tax expenditure instead of income.
This attempt by Chartered Accountant Anil Bokil, must fall into the same category of instant solutions and that itself makes me suspicious. Among the things that it does not address:
1. If transactions below Rs 2000 are to be permitted in cash, what is to prevent everyone from treating every transaction above Rs 2000 as more than one transaction? To give just one example: My wife goes to buy herself a sari which, she is told, is worth Rs 2015/- Under the Bokil law she cannot pay for it in cash. She must pay for it either by cheque or by credit/debit card. What is to stop her paying Rs 20/- in cash and the remaining Rs 1995/- also in cash since now the price of it is below Rs 2000/-? How many billions of such transactions will take place all the time to circumvent the transaction tax?
2. The whole panacea does not take into account the huge unorganized sector. How are the transactions in this sector to be accounted for in the case of all those billions of transactions above Rs 2000? The Bokil panacea seems to have been created by those who live in the cities and are totally unaware of the realities of commerce or what goes for commerce in Kalahandi or Santhal Parganas. Or for that matter in all the Bimaru states,
3. And what about the billions more below Rs 2000. Remember, we are a large country and the villager understands the value of money just as well as his city cousin.
4. The reason for bringing out Rs 500 and Rs 1000 currency notes now has been the inflation. The falling purchasing power of the rupee makes for transactions at an increasingly faster speed to take place in the marketplace. The Rs 1000 note which was demonetized in 1978 has therefore had to be brought back. But that again is taking a different tack. How will demonetising currency notes of smaller denomination help?
5. If the number of transactions have to multiplied in proportion to the amount that is to be spent to avoid tax, this involves that much more effort. Will this not fuel inflation at a runaway best?
6. Since all taxes other than the transaction tax are supposed to be abolished, this will bring into the legitimate marketplace a large amount of spending money, now legal, which will add fire to the inflation.
7. The storm of paper that will result in the banking system will bring it to a grinding halt. Everyone, me and my grandfather included will deal only in cash.
8. The entire monetised economy will come to a halt, with all transactions in cash above Rs 2000 being made illegal.