L I R N E . N E T

LIRNEasia on policy influence and dialogue

Last week, the Sri Lankan government withdrew part of its plan for putting a flat rate tax on SIM cards and further taxing mobile use in the country. LIRNEasia had been concerned about this possible development since it became known. Rohan Samarajiva, Executive Director of LIRNEasia, describes the effort, “with a lot of help from our media friends, and the serendipity of the scheduling of the Broadband Congress for this week that allowed me to hob-nob with the Minister, we rolled back the worst of it.” This happened during a very quick period of six days.

The following documents the different discussions during and after the partial roll-back of the proposed taxes. Please feel welcome to contribute to this dialogue, using the Comment function at the bottom of the page.

Sri Lanka plans to tax mobiles more heavily
14 August

The Sunday Lankadeepa of 12 August 2007 reports that the government has decided to raise the tax on mobile bills from 2.5% to 7.5% and also institute a LKR 50 monthly tax on all SIMs. These are special taxes that are levied over and above the standard VAT of 15%.

Link to LIRNEasia post.
Link to comments on post.

Choices article by Rohan Samarajiva
3 September
Taxing Sri Lanka’s mobile customers; Goose or eggs?

In this article, Samarjiva while defending the need for taxes, uses LIRNEasia research to illustrate why the Sri Lankan government’s proposed plan to levy a flat rate of 50 rupee tax per SIM card at purchase plus a 7.5 percent mobile levy on top of the general taxes would disproportionately affect bottom of the pyramid (BOP) mobile users.

Link to Goose or Eggs article.
Link to comments on article.

Click here to go to LIRNEasia research on teleuse at the bottom of the pyramid (BOP).

Government retreats a little on taxing BOP mobile users
September 6th, 2007

From: LANKA BUSINESS ONLINE - LBO

Sri Lanka has dropped a controversial fixed levy from mobile phones which would have hit the poorest phone users the hardest, but slapped a 7.5 percent tax on calls, telecom minister Rauf Hakeem told parliament Thursday. The government initially proposed a fixed 50 rupee charge which would have hit the poorest or ‘bottom of the pyramid’ users hardest, as well as tripling a usage based charge from 2.5 percent to 7.5 percent.

Link to full LBO article

Click here to read the responses on LIRNEasia
Click here to read the responses on LANKA BUSINESS ONLINE - LBO

Comments from within the LIRNE network….

From Laurent Elder, IDRC
6 September

Great job, Rohan, Harsha and the rest of the LIRNE crew!

Would be interesting though to do some economic modeling to help put
a figure on the “golden eggs” according to the different scenarios…

From Amy Mahan, LIRNE.NET
7 September

Congratulations Rohan. An impressive outcome. The Choices article is very good - and subtle with your goose and golden egg analogy. The government having egg on their face would have been good to work in. [Or not.]

But, I wonder about the punchline to your article that the “fairest way to collect these taxes is by increasing the VAT rate and removing exemptions…” I am not sure what the tax situation is in Sri Lanka, but VAT is usually also seen as regressive. Here (Uruguay is currently undergoing major tax reform) a more fair and bolder proposal is a progressive income tax regime. I guess raising income tax for the wealthy isn’t a popular proposition for many politicians to table.

Also your argument that when hit with higher costs, poor users will make fewer calls, would be interesting to test. BOP studies have indicated that users would increase calling if costs were reduced - but the surveys haven’t asked respondents what they would do if charges were increased. Given that BOP users are already spending a larger portion of family income on communications, one might also speculate that some other critical area of expenditure would be cut, while mobile use remained constant (at the level already deemed necessary).

This is an even more powerful argument - that BOP users are already spending beyond normal percentages for phone use.

From Rohan Samarajiva, LIRNEasia
6 September (moving back and forth through international timezones)

Please see the elasticity discussions at:
http://www.lbo.lk/blogComment.php?newsID=1160489410. I am not sure the elasticity work is complete, but Harsha and Dimuthu can speak to where that is at.

From Professor William H. Melody, LIRNE.NET
7 September

What a great success story, especially since research evidence played such an important role in the debate, actually containing political opportunism.

But this issue is not going away, as other countries are planning similar attempts to capture what they see as “new consumer surplus”, and similar plans can be hatched for VoIP, Internet access and other new technologies and services. If circumstances permit, - financial, skilled staff allocations and other research obligations and priorities - there are synergy gains to be had by pushing this research further, as implied in comments of Laurent and Amy.

Congrats to the research team.

From Randy Spence, IDRC
7 September

Great to see this, and I agree with Bill on the merits of pushing at the boundaries of these issues. On the broad tax issues, public finance theory and experience suggests separating out the progressive income tax as a desirable feature of a (fairly mature) maturing tax system, which can co-exist with indirect/transaction taxes (VAT, sales, import etc), achieve more of the revenue needs over time, and accomplish equity objectives without sacrificing much of the desirable efficiency characteristics of the system.

It is hard to make transactions taxes progressive (or even neutral, given that the poor typically spend more of their incomes on consumption vs saving), but perhaps worth a thought. Exemptions for basic goods and services have often been tried, for example, but unless the basics are consumed disproportionately by the poor, the rich can also benefit (equally or even more). One possibility re mobile use might be to have a somewhat higher VAT rate, but exempt he first ‘x’ rupees from the VAT.. economists may argue against this, as it distorts relative prices and (in this case) may cause disproportionate spending by the poor on mobile communications.. this brings in the elasticity issues.. if the price elasticity of demand is low at the BOP, a ‘tax subsidy’ would not change consumption much, but would provide (progressive) tax relief. The other problem with this kind of approach is administration.. it seems to me it could work for users on monthly plans etc, but hard to see how it could work for cash card users; likely the vast majority at BOP.

Am on a flight, and have yet to read the elasticity material Rohan mentioned, but will look forward to that.

From Rohan Samarajiva, LIRNEasia
8 September

The key question about taxes in countries like Sri Lanka is about least-harm options, not what is theoretically optimal. My line about the increase in VAT has to be seen in context of the repeated statements that reduction of government (and its expenditures) is the best solution.

1. For a population of app. 20 million, LK has 1 million government employees; and half a million government pensioners (former Finance Minister Amunugama’s statement). Paying these people eats up almost all the regular revenues of the government leaving nothing for infrastructure, etc. And this government is hiring more people.

2. Usual mode of operation, particularly of this government, is money printing. Result, we have the highest inflation in South Asia, now running at 17%; was as high as 20% recently. This is an invisible tax that hits the assetless the hardest and actually transfers wealth from them to those with assets (http://www.lbo.lk/fullstory.php?newsID=1510366302&no_view=1&SEARCH_TERM=23). So we have to agree that inflation is bad. It follows then that government must be given money to pay its bills, however unreasonable.

3. All attempts to rationalize an archaic revenue system have been defeated by unions and the people who currently run the government (when they were in opposition). The income tax base is extremely narrow (less than 300,000 income tax files in total for a population of 20 million); the tax rate is 38% and exemptions are almost non-existent (example: contributions to charity are capped at LKR 2500 (USD 25); though there is no limit on what one can donate to government). The process is torturous. People go to all kinds of lengths to avoid opening a tax file, not because they don’t want to pay taxes but because they can’t take the punishment; the arbitrary interpretations, the demands for bribes, etc. If I had any sense I would move from this country simply for this reason. In addition, the government is in dire straits and needs an infusion of money now. No chance that the income tax regime can be fixed in their time frame, other than by jacking up the rate from 38% to maybe 60%. But this is unrealistic and wrong.

4. So what is left is consumption taxes. The readers of LBO know the story upto this point, which is why I start with VAT in the column. What distorts the economy the least is VAT. But the current bunch of politicians in charge have inserted so many exceptions that VAT is yielding less than it ought to. Plus they don’t want to raise the ire of the population by raising the VAT rate.

5. Therefore they are now imposing special taxes. So a tax on the telecom industry, on cars, on whatever. But because these people have an atavistic hatred on the modern the tax is imposed on mobiles, not on fixed. I, and the Minister, both proposed imposing the tax on fixed and mobile both, but Prez refused.

6. With mobiles, they wanted to get an easy-to-collect, high-yield tax so they came up with the fixed LKR 50 tax. This alone would have given them a steady income of more than USD 3.5 million a month (USD 42 million a year). Then on top they wanted 7.5% of all bills (in addition to the 15% VAT).

7. So we went after the LKR 50 fixed tax. They retreated, but raised the levy from 7.5% to 10%. That change causes less damage to all whose bills are less than LKR 2000 a month and increases the burden on those who pay more than LKR 2000/month (pretty much all at LIRNEasia).

Actual amounts will be slightly off because I have not factored in tax on tax.

The significance in the LKR 2000 crossover point can be appreciated by looking at current ARPUs: operator specializing in BOP (prepaid only): LKR 311; largest operator (prepaid LKR 414 and postpaid LKR 1709).

So in discussions of taxation also we must follow the general principle we use in looking at telecom reform: not the theoretically optimal but what is appropriate for the circumstances on the ground. No point in debating the merits of income versus expenditure taxes in the middle of this particular fight. That is a 10-year project.
From Professor William H. Melody, LIRNE.NET
8 September

The issues raised in this discussion remind me of the fundamental steps in achieving positive policy and regulation change that I learned early in my career by bitter experience, and then found support for in the literature - see for eg Chas. Lindblom’s books, Politics and Markets 1977 and Inquiry and Change 1990.

1. stop the proposed harmful changes that are now being proposed;
2. look for incremental steps of positive change in the right direction;
3. develop a good understanding of the long-term goals that you believe are desireable to guide your planning of successive feasible incremental steps toward the “optimum” arrangements.

If you do not succeed in #1, whatever you do on #2 & #3 has no chance of implementation.
If you focus only on #3, you will have no audience outside academia and no influence on policy.

LIRNEasia has succeeded at #1 and the first positive incremental step on #2. The rest of us are providing advice on possibilities for successive incremental steps, and for clarifying the long-term optimal arrangements, both of which clearly need further research as a foundation for further incremental action steps. As I am aware of no developing country that does not tax electronic communication in some way - most of it regressive - there is lots of room here for research-based policy action in other countries, and possible a common research agenda and policy action forum across many countries. The success of LIRNEasia here may provide the catalyst that shows what can be done in this area.

LIRNEasia post:
Effects of telecom tax increases on prices in Sri Lanka discussed
11 September (by Rohan Samarajiva)

Citation from Business article:

Despite increased levy mobile rates unlikely to change
But vehicle prices will go up
By Indika Sakalasooriya
Telecom Analysts are of the view that even though government has imposed a 10 percent tax on usage over the existing 15 percent value added tax, the mobile companies are unlikely to increase prices.
“I don’t think that the four major mobile companies in the country will go for a price hike as there is prospective competition with the arrival of Bharti Airtel by the beginning of the next year. Besides, the industry is growing” said former telecom regulator and industry analyst Professor Rohan Samarajeewa.
He further said that this increase in tax might have a negative impact on people’s usage of mobile phones. “If the levy was imposed the way it was planned, the Rs.50 fixed charge, the impact of it on the industry and the poorer segment of the country would have been horrendous” he said.
“According to one of our studies on the mobile users we have calculated 0.15 elasticity for call charges by the bottom of the pyramid. Therefore if there is an increase in call charges we can assume a certain drop in usage” he further said.

Link to full Business article.

Related posts

• LIRNEasia - hiring in two positions (0)
• Expert Forum on ICT Sector Indicators and Benchmark Regulation (0)
• Diversifying Participation in Network Development (0)
• Mansell Colloquium - Communication and Technology: What’s New? (0)
• LIRNEasia - evidence based policy intervention via the media (0)