## Wednesday, December 21, 2005

### Will BPO hit a staffing crisis?

I wrote a column Will BPO hit a staffing crisis in Business Standard. One of the key reasons why India is doing well today is the revolution in services exports, where white collar staff in India are plugged into globalisation, thanks to improvements in telecom. Today, there are probably a million people working in export-oriented IT and IT-enabled services, giving exports of roughly $15 billion, for an average rate of roughly$15,000 per manyear.

But this sector is experiencing acute problems with recruitment. Employers find that an Indian college education generally induces unacceptable skills, so there is a large mass of unemployable jobseekers, coupled with a shortage of skilled people. Will high growth rates continue to come about, or will India's growth in this area get choked owing to the tiny output of a few good universities?

I argue that India will be able to grow from 1 million to 10 million workers in BPO; from $15 billion in revenues to$150 billion a year. But the rot in the universities hurts at two levels. First, India will be unable to walk up the value chain, to reach higher pricepoints than \$15,000 per manyear. Second, we are likely to experience acute pain, two moves ahead, when the needs of the global economy change, and a large mass of uneducated Indian white collar technicians are unable to learn new skills.

On a related note, in 2001, I had done a Business Standard article on the consequences of globalisation of the labour market for India.

## Sunday, December 18, 2005

### Guidelines for web development

Lots of people I know are involved with making websites, and these websites are often bad. Over the years, I must have written a lot of email criticising this or that.

I finally had a chance to collect together a set of principles that good websites should follow, based on much help from Vikram, Krishna, Sunita and Viral. I'm focusing on websites with lots of content, and not database-backed e-commerce sites, but for this class of problems, these guidelines should be useful. Tell me what should be done better.

## Monday, December 12, 2005

### Updated the China currency regime monitoring page

I just updated the web page Monitoring the Chinese currency regime, using data that goes until 9 December. No, you can't yet reject the null that we're still in the same currency regime as was the case in the first three months (Aug,Sep,Oct). But there does seem to be a fascinating drift in that chart... the next few months could be interesting.

## Saturday, December 10, 2005

### Fun with the Cauchy distribution

Viral showed me a striking result about the distribution of the error introduced in adding up M floating point numbers which are drawn from the normal distribution.

That got us talking about the Cauchy distribution (also known as the t distribution with 1 degree of freedom). This is an infinite variance distribution, so the sample mean is not consistent. It is a striking example to challenge our everyday intuition which thinks the mean is always consistent. So I added an example about the Cauchy distribution to my page of R examples. Try it - it's fun! :-)

## Friday, December 09, 2005

### Oh woe, EPFO

The EPF "rate setting" is a perennial source of conflict and an expenditure of effort in public policy. In today's newspapers, there are atleast four edits on the subject (!): TOI, IE, BS and ET's exuberantly titled Disband the EPFO. And, Ila Patnaik has an excellent, fully reasoned piece titled Subsidy on icecream, anyone?. I agree with everything said in these five pieces :-)

The clarity on EPFO in public discussions - as seen in these five pieces in the newspapers on one day - has come a huge distance. It was not so long ago, in 1995, that the Employee Pension Scheme' (EPS) was introduced. At the time, nobody understood what a mess was being created. It was a certain unthinking acceptance of the notion that pensions are important', and a lack of knowledge amongst local economists (such as myself) about how the field works. Robert Palacios of the World Bank says that in 1995, he was quite clear that EPS had a funding gap on day 1, but I guess at the time, it was hard for people like him and Estelle James to communicate this to the people involved in the decision.

(Though, before we get too optimistic about clarity on EPFO in public discussions', we have to be mindful about the worldview of the hapless lefties).

As with all deep-crises in Indian public policy, the EPFO problem has two layers. The upper layer is the obvious malfunctioning of the organisation: unfunded liabilities, the quest for subsidies, bad administration, potential for fraud, potential for dipping into unclaimed balances, etc. But when an agency malfunctions like this, there is always a deeper story about why the same kinds of employees of government, who are able to produce decent results in some agencies, are producing bad results in this agency. This goes back to the deeper question of legal foundations, governance, mandate, and accountability of EPFO. Solving the EPFO problem is not about minor surgery such as raising the contribution rate of EPS (even though that will help). It is about deeper institutional changes.

Gautam Bhardwaj's team at IIEF has built a great website on this subject including an encyclopaedia on Indian pension reform. My article for the inaugural India-and-China conference of the Lee Kuan Yew School of Public Policy (LKYSPP) at the National University of Singapore is now on my website. It is titled A sustainable and scalable approach to Indian pension reform. It summarises some of the facts and arguments about EPFO, and the rationale of the New Pension System.

### Financial architecture blues

Business Standard has an edit on legislative trouble brewing in Indian finance.

Financial architecture' is a subject of hot discussion in India. (There is a small treatment about this in the paper on Indian finance by Raghuram Rajan and myself in Priya Basu's book). As I see it, one big piece on which there is a clear international consensus is that there should be a lean monetary authority, which only sets the short rate. As Anand Chandavarkar and others have emphasised, it is important to remove fiscal and financial functions from the central bank, so as to free up the central bank to just do good monetary economics, without fear or favour.

This takes you to the second question: To FSA or not? To merge all financial regulation into a single agency, or not? There are arguments for this both ways. Centralisation in an FSA reduces turf conflicts. But in an Indian setting, I can see the dangers of centralisation of too much power into one agency. (I had a piece on this in BS back in 1997, at a gloomy time when SEBI was blocking all kinds of progress in securities).

For a long time, I used to find the FSA solution very attractive. But on balance, I think India will be better off with 4 regulators in finance, to do banking, securities, pensions and insurance. (I had a piece on keeping IRDA and PFRDA distinct, in BS, in 2000). Each of these is a big enough task to merit a meaty organisation of 200-1000 people. It's good to have one CEO of each of these 4 organisations, who lives and dies by the fate of each industry. Yes, there are some messy problems on the interfaces between these, particularly owing to conglomerates like HDFC or ICICI. But having four distinct regulators will also give greater experimentation and innovation. At any one time, one of the four will be a laggard in terms of stifling innovation, but one will be a well-run ship with smart people and doing good things, which will later generate pressure through both precedents and competition upon the other three. If we have just one FSA, it will concentrate too much power into one hand, and the damage that one bad choice of a leader can do is tremendous. Given how badly the appointments process works, it's best to be cautious about this kind of damage.

In this framework, all securities activities would collapse into SEBI, just as is the case with roughly every advanced-country securities regulator. In a typical advanced country, when you go to the futures exchanges, you get to trade - on a single screen - any underlying, ranging from equity to currency to interest rates to commodities to new stuff like weather or catastrophes.

But as the BS edit emphasises, this doesn't seem to be in the pipeline. Instead, we seem to be headed for a bizarre situation with three distinct ponds' : an equity pond, a RBI pond and a commodity futures pond. The exchange infrastructure in each pond will be forced to non-compete with each other. The exchange industry is oligopolistic in the best of times. It's a struggle to get competition into it. Lawmaking like that described in the BS edit will make it much worse.

As the BS edit emphasises, the world is moving towards single-screens that trade everything. But in India, we're forcing the economic agent to make three companies, obtain three or more exchange memberships, have three or more trading screens, and all sorts of compliance overheads. It's inefficient.

## Wednesday, December 07, 2005

### India's new highways

The New York Times has run four articles in four days, by Amy Waldman, on new roads and their impact on India.
Read 'em before NYT makes them unfree! I share their fascination with the transformative power of roads.