First, tech growth is not primarily about domestic per capita income or demand for software in the country. It is about exporting products while also catering to a rising domestic demand. i.e. use domestic demand to find product market fit and then expand globally. So global demand is more relevant than domestic per-capita income.
More importantly manufacturing itself is changing in ways that are lowering capital costs and the barriers to entry into manufacturing. The trend in IoT is an example of that. A lot of new products will primarily be driven by a combination of hardware and software capabilities. Skills that are reasonably widely available in India. (Although China has vastly more talent there.) That positions India in a reasonably good place. So I'm optimistic that domestic tech manufacturing will grow well, but not in the traditional way, such as building semiconductor fabs. It will mostly come from many small companies that make niche-products serving and possibly dominating a narrow global market, with high margins. That's were the employment will come from. An example of this is a company that makes automation kits to convert traditional backhoes into remotely operated machines. The technology requires a combination of software, electronics and mechanical components. And Indian companies (or for that matter companies anywhere) that integrate well into global supply chains for these will do well.
That is simply not true. You could grow anywhere, it's different shades of difficulty though.
> It is about exporting products while also catering to a rising domestic demand.
The revenue from exported domestically made software/tech products, compared to the size of the IT Services industry is miniscule. Almost nothing.
> So I'm optimistic that domestic tech manufacturing will grow well, but not in the traditional way, such as building semiconductor fabs.
As an economy, we should strive for depth - rather than just mate hardware imported from China with some software. Such work is easy to replicate, and is not very different from IT Services in how challenging it can be. Semiconductor fabs however is the kind of thing we should get into.
Basically, an industrial base which manufactures hardware (and other real physical things) is harder to grow than software.
This is a very interesting point. Only recently I had a discussion about this with a colleague.
The painful fact about big salaries in India is either US pays it or a VC pays it. Or you go the services way where some one does the work over the years and has build a company for you, then they pay you services salary(But that is again from the US.).
So its just US or VCs. If you build a company in India and pay you can't pay much.
Concentrating on only one aspect can give answers which we like. Different societies e.g. America, Japan, Russia, UK or China have developed in different ways. I think of 3 basic factors: sophisticated population, lot of land, powerful political system. At least one or combination of more than one is required for rapid development.
India has none of these so it will be painful, long and slow path to development.
The point I'm making is, until manufacturing and industry take off, we won't offset exported IT Services with domestically focused tech of the "software" kind.