New thinking on a traditional public good

Financial Express, 8 August 2009

What should government do? To economists, there is a technical answer: government should raise money through taxes, and spend it on the provision of `public goods'. A public good is non-rival (i.e. the use by one person does not preclude the use by another) and non-excludable (it is not possible to prevent an additional person from benefiting from the public good).

While there are shades of gray in rivalness and excludability, public goods are the zone where government involvement in the economy is legitimate. Protection from war, for example, is a pure public good. When an army is setup which protects the population, it is non-rival (the safety of one person imposes no cost on another) and non-excludable (it is impossible to prevent a newborn child from benefiting from this safety).

An important pure public good is map data. Maps are rival: when I am looking at a map, you can't simultaneously look at the same map. But map data is a public good. If that data is created once and released into the public domain, then myriad private players can use it to create maps, GPS based navigation systems, etc. The job of the government, then, is to run the Survey of India, which is funded by taxes, which creates high quality map data, and releases databases on the website for free download.

Unfortunately, in India, we do everything wrong. Survey of India maps are grossly outdated. On the website, they proudly say: We know every inch of the Nation, because we map every inch of it. However, in good countries, 1:24,000 topo sheets are trustworthy, while Survey of India does not even have good quality 1:250,000 topo sheets. The weakest link about Survey of India is the rules of release. Survey of India is funded by taxpayer money. As a consequence, the information that they create should be freely released back into the public domain for unencumbered use. Instead, Survey of India thinks like a corporation. It has "licensing" restrictions which has effectively made their data unusable.

The most important maps in India today are produced by google. Google maps and google earth are a remarkable combination of satellite imagery and maps, and they are available for free (!). Google has had to reconstruct maps of India from scratch, thanks to the legal problems (and low quality of work) of Survey of India. It is ironic that even though taxpayers are funding Survey of India, this work is useless for the people of India, who are flocking to google maps and google earth. Nokia has also created good maps of India, which are usable through some Nokia handsets (only).

The only flaw with google maps and google earth is that the underlying databases are the private property of google. What would be most desirable is for map data to be a public good, which can be used in all manner of ways by all individuals and companies. As an example, handheld GPS devices (e.g. from firms such as Garmin) are now available for $100. If these are loaded with Indian map data, they can be immensely useful tools for navigation, exploration and business efficiency. Google does not give out their map database to the public, so such applications are infeasible.

Until Survey of India gets its act together, the solution lies with a public domain initiative named `openstreetmap'. This uses Internet-scale collaboration to build maps. It involves volunteers, armed with handheld GPS devices, who are feeding in map data into a central database. This database is a true public good. The licensing conditions of openstreetmap are quite open, though not as open as those used by the US government. Openstreetmap is doing what Survey of India should have done: accumulating high quality map data and releasing it into the (mostly) public domain.

In summary, three strategies are now in play in India:

openstreetmap: A good solution where maps are free and map data is mostly a public good.
google: A good solution where maps are free but map data is unavailable.
Survey of India: A poor solution where maps are not free and map data is mostly unavailable.

The users of maps are flocking to google, Nokia and openstreetmap. The unreconstructed Survey of India is being rendered irrelevant in the field of map data, much as the Department of Posts was rendered irrelevant, a decade ago, in the field of delivery of letters.

What can be done to make Survey of India matter? From the viewpoint of the government, the first best strategy is to shift Survey of India into the mode of uncompromisingly releasing map data into the public domain, matching the release strategy of the US government on openness. Through this, the government would continue to engage in taxpayer-funded efforts at creating map databases, but the full benefits would come back to the people of India. In addition, Survey of India needs to get up to timely 1:24000 coverage of the full country.

If these changes are infeasible, it is better to shut down Survey of India, and transfer its annual budget (of roughly Rs.200 crore a year) to openstreetmap, for the latter is producing public goods while the former is acting like an inefficient corporation.

Back up to Ajay Shah's 2009 media page
Back up to Ajay Shah's home page