O-ring theory is a theory of economic development proposed in 1993 by Michael Kremer. It has been expanded on since then, most famously by Garett Jones, who did work to show how this model could play out on the international stage.
The titular o-ring refers to the faulty o-ring that caused the Challenger shuttle to explode in 1986. The space shuttle was perhaps one of the most technologically advanced machines that the human race had ever produced, with billions of dollars spent its development and construction; but regardless of the expertise put into its construction, it was only as strong as its weakest part, and that single point of failure caused the rest to fail catastrophically.
Kremer proposed that this cautionary tale works as an analogy to explain the way that economies succeed, fail, and self-organize. It states that an economic network is only as strong as its weakest point, and from this, many things follow. As an example of how this plays out in a random industry: good writers prefer to use good editors, and they both prefer to work with good publishers. There is little point in leaving a masterpiece in the care of a drunk editor, or giving your novel over to a publisher that will fail to produce and distribute it effectively. Likewise, a publisher that very effectively distributes crap books will not last long.
Moreover, this applies across collections of industries: a good restaurant needs good farms, good suppliers, a reliable municipal power grid, a good banking system, good culinary school producing good cooks, and an healthy economy producing customers with some cash to spare. Just as a good cook needs a waiter that won't spit in the soup, a good restaurant needs good roads.
This helps explain why poor countries have trouble bootstrapping themselves into rich countries, why rich countries get richer, and why people change cities or countries to work in just the right place. Being at the right place in the right time leads to greater productivity, greater innovation, and most importantly, greater concentration of right-time-right-placeness.
Because having a collection of competent people working in efficient organizations raises the effectiveness of everyone involved, the gains produced by the most well-developed and honed economic systems can dwarf the productivity of the next-best. Even if Annie and Bob are undertaking exactly the same task with exactly the same skill level, if Annie is in a developed economy and Bob is not, Annie will be more productive, simply because she will have fewer unexpected roadblocks to progress.
The flip side is that wages reflect productivity (in theory), and thus there is reason to expect the emergence of a wealthy working class, drawing ever further ahead of the lower classes. Depending on the institutions in place, this may result in a stratified meritocracy, a system of nepotism protecting an oligarchy, geographical wealth disparities, or pretty much everything else we see in our world. Under this model, we should expect inequality to emerge naturally and consistently, and should not expect this problem to solve itself when we have "enough progress"; the drivers of progress are also the drivers of inequality.