The lobbyists themselves state the “Tullock paradox”
According to statistics United Republic assembled, the prescription drug industry spent $116 million lobbying for legislation to prevent Medicare from bargaining down drug prices — legislation that enabled drug companies to make an additional $90 billion annually. That amounts to an extraordinary 77,500 percent return on investment. Oil companies, in turn, had a return on investment of 5,900 percent, and multinational companies, 22,000 percent….
For example, the Carmen Group, a Washington lobbying firm, boasted on its Web site that for every dollar it collected in fees, clients got $100 in benefits.
I enjoyed the ROI visual presented here. Of course the Carmen Group needs to explain why it is not raising its prices and the companies need to explain why they do not spend more. Presumably the rate of return on additional lobbying expenditure is relatively low.
One part of the Tullock puzzle of high upfront but low marginal returns to lobbying may be answered by the article:
I asked a prominent Democratic lobbyist with just over $3 million in annual billings — who requested anonymity to avoid alienating his clients — about the difference between trying to win legislation and trying to block it.
“It’s significantly easier to block and impede,” he said.
Once you have blocked, you have blocked, and “more blocked” doesn’t necessarily make you better off, so you can end up at a discontinuity point. This also may explain why the forces seeking to push new laws do not spend more. Spending more, to try to get your way, may induce more offsetting investment from the “blockers,” who have access to a cheaper technology, so to speak. Still, if they have a fixed cost for getting the blocking activity up and running at all, you may spend some small amount in the first place, enough to have influence but again with the knowledge that their blocking abilities put a discontinuity in the returns function.
The article, with much more useful information, is here, by Thomas Edsall. Here are a few earlier MR posts on the Tullock paradox.
Addendum: Karl Smith comments.