The Obama administration is actually pleading with people to shop around before choosing a health insurance plan. Ordinarily that would be good advice. Ideally, people should comparison shop – considering the differences in premiums, subsidies, and networks of the different plans. But as I wrote at Forbes this week, there’s one big problem: Intelligent comparison shopping is virtually impossible.

The Obama administration calls the health insurance exchanges “markets.” But they are unlike any market you have ever encountered. Think about what you would have to consider to make a reasoned choice. For starters, you have to consider all the health problems you might have. For example, you might get cancer, heart disease, diabetes… Then you would have to compare the networks of all the different plans for cancer care, heart care ….

But wait. It isn’t even that simple. Turns out, cancer isn’t cancer. There are 200 different types of cancer. Cancer can appear in any organ, for example, and there are 60 organs. So you can’t just compare “oncologists” in various networks. You have to consider all the different types of cancer that might afflict you and which network would be best for each one.

Other disease categories are similarly complicated. Apparently there are 100 different diseases of the liver. There are 100 different rheumatic diseases. There may be as many as 100,000 different diseases overall.

Choosing among health insurance plans would be a herculean task, even for the IBM computer Watson. In addition to considering every illness that might afflict you, Watson, of course, would want to know the probability of your getting each one. It might base that judgment on a computer analysis of your genes. But even with that, Watson would need some metric for comparing doctors. That is, knowing that one doctor is better than another is not enough. We need to know how much better in order to make comparisons over thousands of doctors and thousands of diseases. Unfortunately, no such metric exists.

So instead of doing the kind of analysis that would stump even Watson, what are real world people actually doing? They are making choices based on one or two simple parameters and then they are sticking with that choice – come hell or high water. For example, healthy people are looking only at price. People with a serious health problem are trying to find a network that has a doctor they already trust. And that’s about it.

This is what we should have expected all along. James Surowieki writes in the New Yorker:

When the Netherlands introduced managed competition for insurers, in 2006, almost twenty per cent of the insured switched after a year. But, by 2012, less than four per cent did. Karen Lamiraud, an economics professor at ESSEC Business School, studied the insurance market in Switzerland and found that the average switch rate between 1997 and 2007 was three per cent.

Federal employees have been in a similar “managed competition” system for more than three decades – in fact it may be the oldest such system in the world – and the switch rate has been consistently less than 5 percent at annual open enrollment periods.

The experience of private employers has been much the same. Surowieki writes:

Ben Handel, an economist at Berkeley, did a study of the insurance choices made by eighty-five hundred employees at a big company, and found that most ended up staying with their default option, even when it cost them serious money. “Consumers leave thousands of dollars a year on the table because of inertia,” he told me.

So why aren’t these “markets” working like real markets? Because they aren’t real markets. They are artificial markets, created by governments or by employers. They are also replete with perverse incentives both on the buyer and seller side – incentives that undercut all the normal market mechanisms that would tend to solve the kinds of problems we are describing.

So what would a real market look like? I’ve written on that elsewhere, so I won’t elaborate now. But,

• Here is what a real market for health insurance might look like.

• Here is what a health plan in such a market might look like.

• Here is what a real market for medical care might look like.

Finally, on top of all the normal problems you would expect to find in any managed competition system, there is the added problem that Obamacare’s web site is not user friendly. Incredibly so. This is a post that appeared at The Heath Care Blog. I am reprinting it in full:

On the Healthcare.gov web site I was filling the application – an arduous process that – even when pre-filled from last year, takes 30-45 minutes. At the review and sign, I found ONE date that was wrong: the day and month were inadvertently transposed. from 09/08 to 08/09. Since the information will be checked against tax records I thought it best to correct this prior to signing.

I clicked on the “edit” button which brought a box “Do you really want to edit your application”, Yes! That’s why I clicked the button – BOOM! back to “GO”,

So it took almost 45 minutes to go through again, (I do work by the way, so this time consuming process is not OK), but I did it. THEN at review I found I had been so frustrated OR the process accepted the key stroke wrong so I now had 09/03 instead of 09/08.

NOT wanting to go back to the very beginning AGAIN, I called the help desk, thinking this would save time. The agent was supportive and pleasant, but basically REFILLED the ENTIRE form again!!!!!!!

SURELY the US GOVERNMENT can MAKE THIS WEB SITE ALLOW A SINGLE CHANGE OF A SINGLE ITEM. What is wrong with the IT people that I cannot change a single date on this thing without wasting 45 minutes? I guess not. Really? Are these the same people that are in charge of our defense systems?

I have spent about 2 hours trying to correct a single date in the application. I just thought someone might see this and try to make this site better.