Failing to win the contracts after the development work could wipe out smaller organisations so the risk could be along the lines of 50% of winning and 50% chance of annihilation, at which point the sensible course is not to bid at all and find other work to do.
This is common in parts of the UK public sector I am familiar with. Often government departments are forced to require considerable demonstrations as part of the bidding process. This could mean, for example, being forced to produce a publication as part of the bid to get the contract for publications for that department for 2 years. Given the considerable cost of managing a bid, and the considerable cost of producing a demo publication, many smaller firms decide not to try as they know that the odds of winning (when competing with potentially dozens of other firms) are too small to risk the expense.