And if there is no competition, you no longer charge competitive prices, you charge what people are willing to pay. The whole promotional plans they offer are used to figure out how much you are likely to pay. Once the promotion expires and you don't call to get it extended, you are basically saying that you're ok with increase. Once enough people don't do that, the price will go up.
If you compare Australia and New Zealand for instance, there’s some very obvious factors that could explain that difference. Australia has a lot more outbound submarine bandwidth than New Zealand has, with New Zealand also being much further away from the rest of the world. Most of Australia has either no service or terrible service. New Zealand has approx the population of Sydney, while having more than 20x the land area. Australia has about 5x the population of New Zealand, so all of your fixed prices per customer are going to be less in Australia.
For any two countries you’re going to get an incredibly long list of factors that influence price. Population density seems like it should have a significant impact on the cost of building and maintaining mobile infrastructure for example. The US has close to 10x the population density of Canada, and Europe has over 3x the population density of the US, which I would expect is a non-trivial contributor to the reason that Canada is more expensive than the US, which is more expensive than Europe.
EDIT: On the "US is big" point, it's notable that pricing doesn't really follow population density in Europe. The UK has four times the population density of Ireland, but only fractionally cheaper data. Germany has similar population density to the UK, but far more expensive data (though I wonder are they looking at all extant contracts or something; data recently got a lot cheaper in Germany).
All three countries have comparable household earnings. Sweden has a third the population density of Ireland (less than a tenth that of the UK!), slightly higher average data cost, _cheaper_ minimum data cost, and comparable household earnings.
In general, any attempt to explain this via recourse to population density, income or GDP fails in the developed world.