What do you do when you have been given a death sentence? Do you live your remaining time to the fullest or do you spend that time taking long-shot chances at a cure?

No, this is not a column about cancer. It is a column about businesses whose core industries are doomed. The perennial staple of such discussions is the venerable buggy-whip manufacturer, but we could discuss half a dozen more modern equivalents — the makers of camera film, for example, who faced exactly this problem, or Blockbuster video. Choose your disrupted industry and you can find any number of firms flailing, and generally failing.

Take my own business, the media. It is unlikely that in 50 years many people will be getting their news physically printed upon the ground-up carcasses of trees. But at the moment, a lot of people still are. And that has placed the purveyors of news in something of a dilemma: go online and attempt to capture the new market or just keep on keepin’ on?

In a new column for Politico, Jack Shafer suggests that “Go online, young newspaper!” might have been the wrong advice. He bases this conclusion on a new paper from H. Iris Chyi and Ori Tenenboim, of the University of Texas. Their data seems to show that very few American newspapers have managed to capture a significant digital-only audience: The New York Times, The Wall Street Journal, USA Today, the Austin American-Statesman and The Washington Post. In most other places, people like the physical newspaper better. Yet local papers are pouring significant resources into their digital presence — arguably hastening their decline by starving their core product.

Shafer is very sympathetic to this argument. I am perhaps somewhat less so, if only because I view this as a predictable entry to the predictable conversation that you hear about every industry that is being massively disrupted.

At its most basic, this sort of existential crisis presents companies with a very unpalatable choice. In the long run, your revenue will, to a virtual certainty, fall to $0. But in the meantime, you have a cash cow that will still throw off income for quite a while. Do you milk your cows until the last of the herd falls over dead? Or do you take some of your revenue and invest it in a flock of goats, allowing you to pursue exciting new opportunities in the field of goat-cheese manufacturing?

There are always good arguments on both sides. On the one hand, you are probably not very good with goats. You have a large staff of highly competent dairymen and milk maids, who probably are not much interested in goats, don’t have many goat-related skills, and, in part because of the aforementioned lack of interest, may never develop the world-class talents needed to beat out competitors in the cut-throat goat-cheese business.

If you were a brand-new entrepreneur looking to get into goat-related enterprises, you probably would not decide to buy a dairy farm and retool the whole thing for goats; you would probably get some nice hilly land somewhere and populate it with workers who were, so to speak, goat-native. Dairy operations, many economists might argue, should look at the problem as if they were considering entering the marketplace for the first time — or, in econo-jargon: ignore sunk costs. In plain English, this means: don’t try to turn yourself into something you’re not. Milk the cows until they die, then walk away with whatever profits you have stashed in the shed.

These critics will be inevitably vindicated. Most of the dairy farms that attempt the conversion will founder and fail, beaten by newer and, er, nimbler operations that are not weighed down with all their legacy costs.

On the other hand: $0. A 100 per cent chance of $0 in revenue does not necessarily look better than a 5 per cent chance of a profitable goat business; especially if you can turn some of your pasturage over to goats while still getting some milk out of the old cows. Companies that fail to make a change, the visionaries argue, are blinded by their inability to see the market’s future, or by internal resistance from workers and management who just love those cows so darn much.

The visionaries will also, inevitably, be vindicated. A few of the dairy farmers will manage to make the leap, and they will continue on as thriving concerns. Their more — oh, gosh — hidebound competition will inevitably die off. And those deaths will not be as pretty as the sanitised pictures often painted by advocates of the cash-cow strategy. As the old herd dies off, fixed costs for things such as empty stalls balloon in relation to revenue, often bringing the farm’s death much sooner than expected. Workers get landed on the job market with few saleable skills, pensioners who held on to the stock too long have to ride it all the way to the bottom, and the necessary cost-cutting often results in various mishaps that lead the company to leach out its final days amid overgrown pastures and falling-down buildings.

Ultimately, most people who choose either strategy will end up looking like fools. That is not because they are fools; it’s because they are in an impossible situation. They have a lot of valuable skills and capital invested in a market that is dying, and at best those things will transfer imperfectly to newer markets. Whatever they do, you can always present a convincing argument that they have made a bad choice because none of the choices they had were good.

In the meantime, we will have the argument — perennial, inevitable and as unwinnable in media as it is anywhere else. Most of the newspapers in operation at present will ultimately die because the internet rewards scale rather than deep local knowledge. They will die whether they stick to their knitting or go all-in on “digital first”. And their deaths will, as deaths tend to, be rather unpleasant. I’m not going to tell them to waste the time they have left on a long-shot chance at life. But I’m not quite willing to tell them they shouldn’t, either.

Megan McArdle is a Bloomberg View columnist writing on economics, business and public policy