Don't fall prey to the currency conversion fallacy:

Stories on tech workers in India always talk condescendingly about 'ridiculously low' salaries by looking at currency conversion into dollars, without ever considering buying power in the same country...

Converting into dollars is useful when some form of currency arbitrage is going on, earning in one currency and spending in another, but not when both occur within the same country.

Manish -- that assumes that most of the cost is in labor / non-tradeables rather than tradeables.

Yep, the Indian space program tries to do most tech indigenously.

Manish -- and the Indian industry tried to do cars indigenously as well. But the cars they produced were alot more expensive than the ones they could have imported.

On top of which, you have to import some of this -- materials, machines with which to build other machines, specialized parts, etc.

Made in India doesn't necessarily mean cheaper than in the west for a high value added physical product. India does best at bringing down the costs of low quality labor intensive goods (sweatshops) and services. But in the middle, their track record is not so great, especially given the quality problems.

Cars are different than science, where they're expending a lot on salaries. Space sci is more analogous to software development, I think. Perhaps Abhi can jump in here...

Not exactly sure what your precise argument is but space science doesn't really have any equivalent when it comes to R&D and design costs except for major military programs. The industry standard for doing cost estimation is a program called PRICE. This program has a variety of metrics it uses to try and obtain the most likey development and operations costs. The factors applied to anything space related products are always the highest. On top of that, the model does not have enough underlying data to be very accurate and usually ends up underestimating costs. Space is just an expensive thing because of the fact that it is still new. This is mostly independent of what country the work is done in. If for example it was cheaper to make satellites in India we would be doing so (with the exception of military satellites). Since India only has a small fraction of the market it stands to reason that either it is not cheaper or the quality is poor.