Tuesday, April 14, 2009

Problems to think about...

1. Why do we see so many religion based communal riots but not as many caste based riots?

2. Why are we repeatedly getting a hung parliament? 
     Is "hung parliament" a problem? Is there a defect in our democratic system?
If yes, then what is the core issue and how can we correct it? 
If no, then how can a stable government be formed in a hung parliament?
Are such outcomes healthy for a democracy to thrive or is it a symptom of a faulty system? If its is a healthy outcome then how to make a stable gov. while still reaping the benefits of a true democracy albeit with a divided mandate.

3. Is Democracy or the "rule of majority" the best system possible?
.
.
.
Your comments invited...

Procurement of computer peripherals: Leveraging long term contracts

Long term contracts are often used to negotiate cheaper rates from suppliers. Long term contracts can be leveraged effectively but it assumes that the price of commodity/service to be procured will go up or atleast maintain the current value as time progresses. Long term contracts can't be of too much help if the commodity/service price is likely to go down with time. One typical example of such a commodity is computer peripherals which usually show steep fall in prices with time, specially after a period of say 8-12 months after launch. 

In such a situation how should the bidders be asked to quote? It's is almost impossible for even suppliers to intelligently predict the prices of such items 6 months or an year down the line. Under such cirumstances what price should a supplier quote for the items? 

Thus we see that asking bidders to quote prices for a long term contract may not be the best option as nobody has enough knowledge of how the prices are going to behave after say 6 to 8 months or an year. Computer peripherals show steep decline after certain period which typicalls depends on the product and the kind of innovation we are seeing in that product line. If we ask suppliers to quote their prices they will base their quotation on present prices which might be much higher than what the prices will be 1 year later. Then how to leverage a long term contract in such a case?

What should then we ask for, from bidders, if not prices? How about asking them to submit the percentage discount on MRP (Maximum Retail Price) that they want to offer for a long term contract (where MRP keeps changing with time). A further extension can be to link the discpunt relative to an index for the good/material or its cost driver. I believe that suppliers are better positioned to quote a discount (and thus keeping in mind a profit margin) that they can offer rather than an actual price which will be fixed over the period of contract. 
We can then compare bidders on the value of discount quoted.

How to Monitor Revenue being earned at Far Away Kiosks?

There are many public/private initiatives where performance has to be monitored in far flung areas which are difficult to approach let alone monitor. We faced a similar situation in one of the places where I was working. Our project was to successfully run over 2000 ICT enabled kiosks in rural areas of Himachal Pradesh. These kiosks were supposed to be run by local level entrepreneurs and we were responsible for setting up the infrastructure and making various services available at the kiosk. The kiosks were set up in one of the most inaccessible terrains of Himalayas and of course had its own set of difficulties. Kiosks were there to disseminate different kind of G2C as well as B2C services to their respective panchayat (a collection of around 6 villages) population.  Thus each panchayat was supposed to have one such kiosk.

Now let's understand what exactly were we trying to monitor. We wanted to monitor 
1. What kind of services are being used in each panchayat. 
2. How much is the kiosk operator earning from each service. 
Services included printing, photocopy, Internet access, fax, online booking of railway/bus/airline tickets, e-education, Gov. forms (like birth/death certificate, land records, etc.), et al.  Some of these services were provided completely online through a portal provided by us and there were appropriate provisions to monitor their usage. However, many other facilities like printing, photocopy, fax, scanning, selling of gov. forms, etc were few of the services which were extremely difficult to monitor. We could have never found out how much is the operator earning through these services and they incidentally were supposed to be the most popular services. 

Now how can we measure the usage of such services as accurately as possible to gauge the actual earnings from each kiosk? Of course here is an assumption that the operators will understate their earnings and their are plethora of reasons to support that but I will not go into details discussing them in this post.

Well, to take care of the revenue earned through printing, photocopy, scanning, internet access, et al can be monitored through technology. For example, there are Photocopy/Printing machines available which keeps a count of number of printed/photocopied pages, etc. But the problems were
1. Who will go to each centre to note down the readings? Even if machines are intelligent enough to automatically mail the readings online, still the function can be easily manipulated or better put the machine offline while using it. 
2. These machines were way over our budget.

So what else can we do? Install cameras and monitor what is happening in each kiosk? Nope, surely an impractical idea.
This is the plan I proposed. Well not 100% secure but it uses a combination of incentives and punishments to get as close as possible to our goal. 
We can use two approaches in tandem.
1. We know that each kiosk has a different set of population to cater to and not all kiosks can earn similar revenue. We planned to divide all the panchyats on the basis of per capita income and population. Thus in a way we tried to measure the earning potential of each kiosk. Based on this data we made multiple categories and grouped together kiosks with similar earning potential. Our idea was to ask each kiosk operator to declare their earnings through each service themselves. Of course in the absence of any tangible monitoring system specially in case of print, photocopy, internet access, scan, selling of gov. forms, etc the operators were free to declare a deflated earning. So we brought in a clause of performance monitoring which stated that if an operator's declared monthly earning goes "below 20% of the average earning stated by the operators of his category" then he will be put in the warning zone and he will have to improve performance or be ready for eviction. In the absence of knowledge about the kiosks which fall in each category, its difficult to collude and declare revenue figures close to each other every month to evade falling in the eviction zone.
2. Introduce value cards for every customer of the kiosks and allow the card to be used across over 2000 kiosks all over the state. Inform every customer that every transaction done, no matter how small, adds up value points on their card and the points can be redeemed for cash, gifts, etc. This gives the incentive to the customer to feed in the information about each and every transaction against their value card. This will urge customers to ensure that all their spending in the kiosks are fed into the system by the operator. 

Thus a simultaneous implementation of the above options brought us close to estimating the true business being done by each kiosk.

Saturday, April 4, 2009

"New bill to curb traffic violations": Delhi Gov. Will it Work?

How to effectively implement traffic laws?

There is no denying the fact that violation of traffic laws is rampant across India. Well, Delhi has its own share of problems too and Delhi Government looks determined to put things right this time. To achieve the same, the newly elected Delhi Gov. has come up with the plan of hiking penalty charges for breaking traffic laws. For instance, the new Bill proposes to hike the fine for jumping red lights or not wearing seatbelt from Rs 100 to Rs 500. It is also proposed that parking in no-parking zones would invite a fine of Rs 1,500 instead of Rs 100 at present and so forth.

So how much is this going to help?

To find out if this is an effective solution we first need to understand the problem in detail. The problem is not only a lenient system which doesnt adequately penalize the violators but also an ineffective law enforcement agency (traffic police). No matter what happens on road, an offender knows that he can easily bribe the traffic police personnel and get away. The problem is that by hiking the penalty we are giving even more reasons to the offender to bribe the policeman rather than paying up the fine. For a fine of Rs. 100 an offender used to pay Rs 10-20 to get off, for a fine of Rs 500 he will now perhaps offer Rs 50-60. Surely the offender would try even harder to evade the hiked penalty. Under such circumstances how can we ensure proper implementation of traffic laws? Surely an increased penalty would also mean an increased bribe and a bigger dent in offender’s pocket and thus a bigger deterrent. However, this is surely not enough. It would never be able to bring desired results.

What incentive can we give an offender so that he prefers to get punished rather than bribing? I don’t see any particular way of doing that until and unless penalties are made too lenient which is surely unacceptable. Thus offenders will always find a way to escape fine by bribing policemen till policemen are ready to accept the gift.

Well, then can we influence the policemen to reject the illicit offers? Of course one option is to use technology to better monitor whatever is happening on the road as well with the policemen. However technology is expensive and its implementation will surely take time. So let’s forget technology for now. Well, just like we have heavy penalties for traffic rule violators, similarly we can have even heavier penalties for policemen found guilty of taking bribe. Also make sure that penalties are not only monetary but also career threatening for greater impact. But who will catch the corrupt policemen? Non-Gov. agencies can keep a vigil but still not an easy solution to implement. Another option can be to incentivise giving a ticket, say through sharing a portion of revenue earned through penalty with the police personnel. But this method can’t be overstretched as it might lure policemen into penalizing innocents for their own benefit.

So we have tried heavily penalizing offenders for breaking traffic laws and also traffic policemen for accepting bribe. However, this still doesn’t seem to be enough. Can we also introduce a system which ensures more cases to be booked rather than let off through bribes? I believe there lies another option in numbers and statistics. How about finding out the average number of violations, in say Delhi, each day? Can we break down the result in the type of violation, area in which committed and time during the day when rules were broken? There are numerous NGOs, NPOs and even Bhagidari which can collect this impartial data specially related to seatbelt, jumping red lights, rash driving, parking in no parking zones, use of mobile phone while driving, et al. It is for everybody’s benefit to regularly monitor this data. To begin with, we can take an average number of violations of each type, committed on each day in each area. Then we can ask the law enforcement agency to at least ensure that we bring to books at least 60% of the violations happening each day in the respective areas. This figure can be increased gradually. This will not only give invaluable regular inputs about traffic law violations happening across the city but also ensure that majority of them are brought to books rather than ending up in bribes.

Of course even the combination of all three options pointed out above can’t ensure a perfect system but it can surely be a step towards an improved system with lesser loopholes.