In the US, solar power remains one of the pricier options; without tax breaks and renewable energy mandates, we’d probably see only a fraction of the installations that are currently being built. And, without much in the way of on-grid storage, solar is mostly useful for cutting into the peak demand that tends to hit mid-day.
All of which would seem to indicate that there’s even less of a reason to install solar outside of wealthy, industrialized nations like the US and Germany. After all, if it doesn’t always make economic sense here, it would make absolutely no sense to install it in developing nations, where the cost matters much more and getting reliable power is a real challenge. But that reasoning may very well be wrong; the things that are problems in the US may not apply at all in countries that don’t have an existing grid or the large generating facilities that are needed to feed power into it.
Coal generating facilities only become economic and efficient when they’re relatively large. Natural gas plants can reach pretty good efficiencies when they’re a bit smaller, but tend to need a large infrastructure to feed them with fuel. These factors mean that fossil fuel generating facilities make the most sense when plugged into a grid, which can then distribute their output to an appropriate number of homes or facilities.
Of course, the grid itself costs lots of money to build and maintain. And, to function effectively, it has to be stable: supply and demand have to be matched, the grid frequency has to be stabilized, etc. This tends to be easier to do with scale, when you have multiple plants feeding in to the grid, so that if one fails or is taken down for maintenance, the rest pick up the slack to stabilize the grid.
All of which means that building a grid is expensive, and maintaining it involves significant ongoing investments. Bringing electricity to new areas isn’t as simple as paying the price for a single fossil fuel plant. And the cost per capita goes up by a lot if the population being served is a diffuse, rural one. (Meaning one that needs many more kilometers of wiring,)
In these contexts, solar can make a lot more sense. The generating capacity can be placed right where it’s needed, minimizing the need for infrastructure and lowering maintenance costs. And the power can do useful things despite its intermittency: pump water that can be purified and stored, charge devices for use later, run equipment in small workshops, etc.. It’s not a perfect solution, but it’s one that can be put in place now at a moderate cost, rather than at some indefinite point in the future as part of a large, multi-year infrastructure project.
This may also be true in areas that are already served by a grid. In many locations around the world that have been wired up, keeping the grid stable has remained a challenge. As a result, outages can strike at random, and last an indeterminate amount of time. In many cases, it might be preferable to have some times of the day where it would be guaranteed that power would be available through something akin to a micro-grid.
(Wind fits into this mix somewhat awkwardly. The most efficient wind generators are large, cost more to maintain than solar, and don’t generate power on any predictable schedule, which tends to make wind more appropriate for grid-0scale generation.)
There are a lot of other factors that come into play. In most cases, projects won’t involve building a grid from scratch, but rather extending or stabilizing an existing one. And there are factors beyond infrastructure costs to consider. Industrialized nations have taken a century to make progress on things like the pollution lofted into the atmosphere or left behind after the burning of coal. Doing so adds costs, as does failing to do so (as China’s air quality is now ably demonstrating).
All of which is to say that the right answer will be complicated, involving compromises and changing from one situation to the next. But it’s precisely the fact that it’s complicated that gets ignored completely by the view that solar is too expensive for consideration in the developing world.