Krugman Too
In today’s NY Times Krugman, in passing, obliquely referenced the results I described yesterday, that Medicare beneficiaries are more satisfied with Medicare than are the non-elderly with private insurance.
Today, Medicare — which is, by the way, one of those “single payer” systems conservatives love to demonize — covers everyone 65 and older. And surveys show that Medicare recipients are much more satisfied with their coverage than Americans with private insurance. [Emphasis mine.]
Despite what Krugman and others say, these results do not show and cannot be interpreted as showing that public insurance causes greater satisfaction than private insurance. Such results, biased by sample selection, are handy rhetorical tools, but they don’t actually tell you anything.
(Actually, little secret, those survey results also show that Medicare beneficiaries enrolled in Medicare Advantage (private plans operating under Medicare) are a bit more satisfied than Medicare beneficiaries enrolled in traditional fee for service Medicare. There are selection problems with this comparison too so I won’t conclude that it means that people like private plans more than government ones. Would Krugman?)
Profiles in Selection Bias
It isn’t hard to find selection bias once you know how to spot it. Yet many people, including reporters and bloggers, don’t see or willfully overlook it, even in its simplest form. It is so much fun (even profitable) to report results that suffer from selection bias because they “show” such “interesting” things. I think the media actually loves selection bias and doesn’t know it.
Quick review: What is selection bias? Suppose a study finds that a group of individuals with larger feet have higher on IQs than a group with smaller feet. Does having big feet cause people to be smarter? No. Big footed people are older and have more education. The selection criterion (foot size) for the groups across which IQ is compared leads to a bias in the estimate of the causal relationship between foot size and IQ because the two groups differ in systematic ways that are important (age, education). Clearly taking growth hormone to increase one’s foot size is unlikely to make one smarter.
Recent posts by Kevin Drum (Mother Jones), Mark Blumenthal (National Journal), Matt Yglesias, and Ezra Klein (Washington Post) all draw conclusions from comparisons that potentially suffer from selection bias. The first three report the same comparison of survey results on the satisfaction of individuals with their health plan. Medicare is associated with greater satisfaction among its beneficiaries than is private insurance among its policyholders. Drum, Blumenthal, and Yglesias claim this indicates that people really like (or would like) a government run health plan better than those provided by private insurers. They suggest that the public nature of Medicare causes greater satisfaction. (By the way, Krugman has also implicitly made this argument too.)
That may be so, but these comparisons don’t illustrate that. The samples for which satisfaction is compiled for the different plan types are systematically different in ways that could bias the findings. The Medicare population is very different from the population that has private coverage. The most obvious way is in age, but also in health, income, and others. The documentation that accompanies the results upon which these statistics are extracted illustrates other ways in which the samples differ.
Ezra Klein’s selection bias problem arises in comparing survey results from populations of different nations on the proportion that thinks their national health system should be completely rebuilt. Do differences in the characteristics of the health systems of nations cause the differences in population level of dissatisfaction with those systems? Perhaps, but one can’t conclude it from the simple comparison that Klein reports. Not only do the health systems differ but so do the populations. Which is the cause of the differences in satisfaction levels? One can’t tell. It is the same problem as above: comparison across samples with systematic differences can produce biased results.
There is no way to know the extent of bias in studies based on non-randomized samples (such as those above) without using more advanced statistical methods (which, by the way, is what some health economists (and others) do–ahem.) Therefore, by themselves the results shown by Drum, Blumenthal, Yglesias, and Klein referenced above are useless in supporting their arguments. Sadly, reporting of results that suffer from selection bias is very common. They’re in the news every day and they’re meaningless.
When To Enroll in Medicare Part D
This post has been cited in the Money Hacks Carnival #76, hosted by Moolanomy.
It isn’t so easy to figure out the right time enroll in a prescription drug plan under Medicare. If you’re healthy you may not need drug coverage. But if you delay you will incur late enrollment fees. What’s the optimum time?
First, some background. Medicare Part D is the component of the program through which Medicare beneficiaries purchase prescription drug coverage. Most beneficiaries who obtain drug coverage through Part D do so by enrolling in a stand-alone prescription drug plan (PDP). Many beneficiaries who do not enroll in Part D have creditable coverage from other sources (a former employer, the VA, etc.).
When the legislation authorizing Part D was under consideration some thought that only beneficiaries who expected to have high drug costs would enroll. That is, it was thought that PDPs would experience high adverse selection, that they would be like “providing insurance for haircuts.” Adverse selection severe enough can cause an insurance plan to die by requiring it to charge higher and higher premiums as relatively lower risk individuals drop out (the so-called “death spiral“).
To reduce adverse selection beneficiaries are encouraged to enroll in a Part D plan while they are young and relatively healthy. The chief mechanism that encourages early enrollment is a financial penalty in the amount of 1% of premium per month that enrollment is later than one’s initial eligibility (penalty waived in the case of creditable coverage).
This presents an optimization problem. Clearly if you buy drug coverage when you’re healthy and use few drugs you spend a lot on premiums for little benefit. On the other hand, the longer you wait the enrollment penalty grows.
A recent paper by Atherly and Dowd in Health Economics (Should Healthy Medicare Beneficiaries Postpone Enrollment in Medicare Part D? 18(8), 2009) concludes that purchasing Part D drug benefits when first eligible to do so is optimal, on average. Beneficiaries should not wait, even if they’re healthy and don’t use drugs. (This does not apply to individuals with creditable coverage for drugs outside of Part D.)
Because the Part D program is only a few years old and beneficiary-level data pertaining to it is not yet available to researchers, Atherly and Dowd took a simulation approach. Using Medicare Current Beneficiary Survey (MCBS) data they estimated by age and gender the probability that a healthy beneficiary will contract a health condition requiring drugs and the probability that individual will die. They then used Monte Carlo simulations to estimate expected lifetime drug spending with and without Part D coverage. Postponing Part D enrollment until one contracts a drug-intensive illness results in lifetime expenditures that are 10% higher for women and 6.5% higher for men, on average.
Thus, on average it is better to purchase Part D coverage as early as possible, though not that much better. If you think you’re likely to be healthier than the average beneficiary you might just beat that 10% (for women) or 6.5% (for men) spread. Is it worth the risk? Likely not. I would perfer the peace of mind of the coverage and use my mental energy to enjoy my retirement.
Who Has Market Power?: Health Insurers vs. Providers
This post has been cited in the 5 August 2009 Health Wonk Review, hosted by Disease Management Care Blog.
Several recent reports and journal articles describe considerable market concentration among health insurers. And there has been debate in the blogosphere about the merits of competition among insurers. Should consumers worry that increased market power by insurers means higher premiums than would otherwise exist in a more competitive market? Or should consumers rejoice that the buying power of a larger insurer will command lower prices from providers thereby keeping health care costs in check?
These are tricky questions and ones I’ve raised before. But now I have some answers from–wait for it–health economists (naturally).
In his 1998 Health Services Research article Managed Care, Market Power, and Monopsony, Mark Pauly lays out some helpful theory. In the case of a monopsony insurer (a market with only one health insurance plan) overall welfare (consumer plus producer surplus) is lower than in the case of a more competitive market. However consumer surplus (the value of the product to consumers less the price they pay) may increase depending on the type of insurer. A for-profit monopsony health insurer may not pass the lower provider prices to consumers through lower premiums. A nonprofit monopsony health insurer, on the other hand, will (in theory).
OK, what does this mean? It means that, according to theory, consumers get the best deal when the health insurer has considerable market power (monopsony or market share concentrated in very few insurers) and when the insurer is a nonprofit entity (as would be the co-ops recently proposed by Senator Kent Conrad). Nevertheless, a monopsony insurer reduces producer surplus (and therefore overall welfare) by extracting prices from providers below those of a competitive market.
If you care about the welfare of producers (doctors, hospitals), as the American Medical Association (AMA) does then this is a concern. To what extent should consumers care about the welfare of producers? Perhaps it is relevant to know whether or not the producer market is competitive or monopolistic? If producers are monopolistic and extracting extra profit (rents) then there is justification in breaking them up, doing so would be welfare improving. If the provider market is already competitive then perhaps providers would be unfairly taken advantage of via a monopsony insurer.
This suggests two ways to achieve lower health care prices: (1) support a trend toward monopsony in the insurer market (welfare reducing though possibly consumer welfare improving, but not always) or (2) break up monopolistic providers (welfare improving). Both would lower prices, and they are not mutually exclusive. If a policy is implemented and we observe a reduction in prices is it welfare improving or not? The answer hinges on distinguishing between cases (1) and (2). Pauly explains that if the quantity of care declines along with price then that signals a trend toward monopsony insurers. A monopsonistic insurer achieves lower prices by holding down demand. On the other hand, if the quantity of care does not decline with price then that signals a trend away from monopolistic providers.
At least two papers have used the theory expressed by Pauly to test whether insurer market concentration reflects monopsony power or monopoly-busting power. In Do HMOs Have Monopsony Power? (International Journal of Health Care Finance and Economics 1(1), 2001) Feldman and Wholey found that HMOs used their market clout to offset the monopoly power of hospitals. They also found no evidence of monopsonistic behavior with respect to physician services.
The findings of Bates and Santerre (Do Health Insurers Possess Monopsony Power in the Hospital Services Industry? International Journal of Health Care Finance and Economics 8(1), 2008 [free working paper version]) are consistent with those of Feldman and Wholey. Greater health insurer market concentration is not associated with monopsony power and suggests that insurers use their power to offset monopolistic providers.
Taken together, the articles reviewed above indicate that insurer market power should not be the focus of attention. Instead, provider (hospital) market consolidation ought to be more closely examined. I would be overstepping my knowledge of the literature to say that antitrust action against large provider groups is the key to more competitive health care markets, but the papers I described above are consistent with that idea. But what of the consolidation implied by the notion of accountable care organizations (ACOs)? I raised this question in a prior post. Some commentators have already pointed out that ACO-type incentives have led to provider consolidation. I hope policymakers are paying attention.
On the other hand, how do you think Democrats will get/keep provider organizations on board? Not by threatening to break them up, that’s for sure.
Budget Tracking and Projections (with Quicken Tricks)
This post has been cited in the Carnival of Personal Finance published on 3 August 2009.
This is the third in a series of posts on investment planning. For those who haven’t read that first post (or have forgotten), I’m soliciting feedback (tips, tricks, links, etc.) that I will cite and use in the final post of the series. Here’s a list of the other posts (numbers 4-8 to appear in subsequent weeks):
- Investment Planning: The Series
- Household Budgeting the Easy Way
- Budget Tracking and Projections (with Quicken Tricks) [this post]
- Willingness, Ability, and Need
- Estimating a Retirement Budget
- Need for Risk: The Details
- Multi-Period Planning and Asset Allocation
- Investment Planning: Reader Tips, Tricks, and Links
I assume all readers of this post have constructed a descriptive household budget with positive cash flow (if not, see the prior post). The bottom line of such a budget is a monthly (or other period) average net income. That is, it is income less all necessary, regular expenses. For simplicity, I will call this amount your surplus and I will assume it is a monthly value (this is not important; if you prefer a yearly value, that’s fine).
Your surplus represents how much you can spend on extra stuff or invest per month. Therefore, if you have a goal, like saving $20,000 for a new car, you can use your surplus value to see how many months it will take you to do so. If you have multiple goals you can set up a spreadsheet to allocate fractions of your surplus to each of them and then compute how long it will take to reach those goals (or vice versa: set the time by which you need to reach the goal and compute what fraction of surplus is required). These examples illustrate that surplus is the key to investment planning. I’ll have a lot more to say about that in subsequent posts. First, let’s apply our budget to the fundamental problems of budget tracking and projection.
I use Quicken to track and project my budget, but you can do all of the following with other software. (Until five or so years ago I used Excel for these functions exclusively. Now I use Quicken 2007.) Since I hate to read software guides I’ve made this entire approach up myself. It would not surprise me if there were other ways to do the following in Quicken (feel free to suggest them or provide links to such things).
I use Quicken as an electronic checkbook. Since I write checks or have electronic payments drawn on and deposited into two accounts, a checking account with Bank of America and a taxable money market account at Vanguard, those are the only two real-world accounts I bother to track (i.e. download/reconcile) in Quicken. Thus, the sum of values of those two accounts is all that is relevant for tracking actual net household cash flow. Let’s call this sum my actual balance. All the other investments I have are not relevant and we can ignore them for this post.
In addition to these two accounts in Quicken, I have set up a third, which I will call my projected balance. This projected balance account is fictitious in that it does not correspond to any real-world account. Using Quicken’s functionality to schedule automatic bills and deposits I have translated my budget (see spreadsheet example) into automatically recurring Quicken transactions into and out of my projected balance account. I have set the timing to correspond to when such payments are actually made in the real world.
At any point in time I can compare my actual balance with my projected balance. The former reflects my real-world income and payments. The latter reflects my budget. The two should match, or at least be close. If they are not, it indicates a problem (e.g. bank error) or, more likely, that I forgot about a big purchase or a large financial gift. When the two are off by more than about $1,000 I search my brain and records for the source of the discrepancy. This doesn’t happen often and when it does it doesn’t take much time to find the source. Then I adjust my projected balance accordingly to bring it back in line with reality.
This real-time budget tracking is very handy. I can always tell if my household is overspending or if we’re saving at a greater rate. This really puts my budget to work.
Another way to use Quicken to put your budget to work is to use it for projections. The “Overview” tab of my projected balance account has an “Account Balance” graph with an “Options” menu. One of the options is to “Forecast my future account balances.” By entering my budget into this tool I can immediately see my projected surplus growth path. It also automatically calculates the monthly surplus (Quicken calls it “net savings”). In fact, I don’t use a spreadsheet for budget projections anymore. I only made this one for illustrative purposes for this investment planning series of posts. It is based on the information I’ve already entered into Quicken. (One gripe I have with Quicken is that it doesn’t seem to be able to base its forecast on the transactions I’ve already entered in my projected balance account. I have to reenter them in the forecasting tool. If anyone knows how to connect the two explicitly please let me know.)
As I said, one can do all of the above with other tools. Being able to reconcile one’s real-world cash flow with one’s budget and to forecast surplus are worthwhile capabilities. In subsequent posts in this series I will show how to use these capabilities in the service of investment planning.
Blogosphere Ethics
If you overhear a remark made by a friend that others you care about (members of your family, say) interpret as racist, what should you do? Do you have a moral obligation to speak out?
Things like this make me squirm. I had trouble finding good guidance on the internet. (The NPR ethicist is somewhat vague on this point. Come on!) Since I would like less racism to exist in the world I would be motivated to take some sort of action, but what, and how?
Does it matter what my relationship with the speaker is? As a practical matter it does. While I may have a moral obligation to do or say something, I also have obligations to keep my job, to keep my family safe, and so forth. Thus, if the speaker is my powerful and resentful boss (in actually my boss is no such thing) I might approach the situation with great caution. If the speaker is a violent thug I would walk away. If the speaker is a friend I would ask her to reconsider her words, to reflect on their impact on others, to contemplate whether they need to be said even if thought.
What about in cyberspace, where one can have relationships without meeting? What if the speaker is such a cyber-friend? Online one can link, comment, and cross-post in ways that suggest endorsement. What if the words were not overheard, but read on a blog, one to which you have contributed? What are the ethics in the blogosphere?
This situation has come up and now I have to figure out what to do. I am not going to publicly name the blogger whose post my family and I find offensive. I do not think that is constructive or important. I did, however, communicate with him privately to suggest he consider removing the post because it could (did) cause offense, reflects poorly on his site, and perhaps those that advertise, post, or link to it. I’ve removed all such links from my site.
I love the internet. I love that anybody can express anything they like. I respect everyone’s right to free expression. But I don’t want to be associated with some of it. The blogosphere is a strange, wonderful, and interesting world. I’m still learning how to behave in it. Perhaps we all are (?). I’ve published this post because I find online relationships, communities, cultures, and ethics interesting, if not important. I welcome comments, thoughts, and links about such matters. If you’ve got any, lay them on me.
Client L, Part 2: Roughing It Out
I’m meeting weekly with a friend, "L," to guide him through the retirement investment planning process. (All my posts about my meetings with L are tagged "Client L".) At our first meeting we drew up a roadmap to guide our planning.
His homework for the first week was to start to co-locate funds at Fidelity and to play around with the retirement calculators on the Bogleheads Wiki. L made some progress on these assignments but didn’t complete them (that’s OK, I’m not in any hurry). Between our first and second meeting he obtained the paperwork to move funds to Fidelity. Before completing them he wanted to review with me the tax implications of moving funds from a former employer’s 401(k) to a rollover IRA. I explained at our second meeting that such a move is not a taxable event so he has nothing to worry about.
At our second meeting we also talked through how to rough out how much one should save for retirement. This is not a replacement for careful investment planning (see my investment planning series for that), but it is a useful place to start. Here’s what I explained:
In general, though not universally, one will need in retirement something like 50% to 70% of one’s current income. For simplicity, let’s take the number 60%. Thus, R = 0.6C where R is retirement income and C is current income (all in constant dollars). The amount of money one can safely withdraw from one’s retirement nest egg annually is, relatively conservatively, between 2% and 4% of the initial principal. Let’s take the number 3% as the safe withdrawal rate. Thus, R = 0.03P where P is one’s retirement nest egg on the day of retirement.
Using these two expressions for R (retirement income) we can solve for P (nest egg) in terms of C (current income): P = 20C. That is, a rough approximation of what one needs to save for retirement is 20 times one’s current income. If your current income is $100,000 then you need $2 million saved for retirement (in current dollars). Don’t panic! This is a rough estimate. Don’t rejoice! This is a rough estimate. It is possible for it to be off by a factor of two. One must do more careful budgeting and planning to obtain a more precise estimate.
L and I also played around with the Ballpark Estimate calculator. We were pleased to discover that he seems to be on track for retirement after all. He just needs to keep it up and to organize his finances so he can manage and track progress.
L’s homework for our next meeting is to complete the movement of funds to Fidelity and to draw up a current household budget, as described in Household Budgeting the Easy Way. That should enable us to get a much better sense of his ability to save and will serve as the basis to determine his needed retirement income. If he has more time I also suggested he open a Roth IRA at Fidelity for himself and his wife. Also, he might begin to find information about Fidelity’s funds and those available through his current employer’s 401(k), particularly their lifecycle and index funds. We’ll be needing to know his options after another meeting or two.
From There to Here, from Here to There Economics Is Everywhere
A recent paper by Miller and Watts explores the economics content of children’s books written by Theodor Geisel, a.k.a Dr. Seuss, Theo LeSieg, and Roseta Stone (Oh, the Economics You’ll Find in Dr. Seuss! (pdf)). (Hat tip to Catherine Rampell of the NY Times’ Economix blog for bringing it to my attention.)
As you might imagine, the paper is eminently readable. Among its contributions is a table listing all the books published under all of Geisel’s pen names and the economics concepts Miller and Watts argue they illustrate. The text makes the case in greater detail, citing plot lines and passages and linking them to economics.
For instance, Miller and Watts cite as an example of asymmetric information the fact that the Cat (in the hat) knew he could clean up his playful mess rapidly using special technology but the kids did not.
Well, there are lots of good tricks one can do with Cat in the Hat and other Seuss books. Miller’s and Watts’ trick is cute, but I’m not sure I buy that there’s much economics in an idea if there is no economic consequence. For instance, there is asymmetric information everywhere, in nearly every story. Does that make all such stories relevant to economics? I don’t think so. Nevertheless, I applaud Miller and Watts for a fun piece of work. Perhaps there is a little (just a little) economics in some of Seuss’s stories. Anyway, it’s a neat way to think about the concepts.
But I Don’t Want the Ribbon: Microsoft’s $4B Waste of Time
My employer has switched to the Office 2007 suite so I am forced to confront the ribbon, which replaces the toolbar menus with which I had been familiar with an explosion of icons. Like Kramer from Seinfeld (video below), I’m not interested in the ribbon. I don’t want the ribbon. I hate the ribbon. Yet, I’m stuck with the ribbon. Also like Kramer, I’ve been beaten into submission. I’ve been abused by Microsoft, again.
Since I cannot revert to the old look and feel I must spend time relearning how to do in Word, Excel, and other Office products what I once did quite well. How many others are wasting time relearning how to use Office software? How much in lost productivity are we paying in addition to the price of the software?
The U.S. labor force has about 155 million workers that use something on the order of 100 million PCs (not every worker uses a PC). For an order-of-magnitude estimate, let’s say they’re all using Microsoft Office. Let’s assume they all will have to convert to ribbon-based Office 2007 or a similar version. Let’s suppose each PC has only one user and that each user will spend about one hour regaining the level of productivity (s)he had with the old, pre-ribbon version of Office. That’s something like 100 million wasted person-hours. A reasonable guesstimate of an average total compensation of a U.S. office worker is perhaps $40 per hour (roughly half in wages, half in other benefits, say). So, workers might waste about $4 billion because of the Microsoft Office ribbon. (One can argue about each of these input numbers but I think any reasonable estimates will yield an order-of-magnitude figure in the single-digit billions.)
All that lost productivity and for what? Are there compensating efficiencies to be gained by using the ribbon as opposed to the old menus? I’ll let someone else make the argument but let’s just say I’m skeptical. If I find any new killer apps embedded in the new Office I’ll let you know.
The Twelve Balls Problem
Alex Tabarrok proposed a problem similar to one proposed to me in college. You can read his problem on his blog. Below I state mine, which you can also find elsewhere on the internet. I kept my notes on how to solve it which I also include below (don’t peek if you want to try to solve the problem on your own).
Problem: There are 12 balls, one of which is a different weight than the other equally-weighted eleven balls. Using a pan balance three times find the anomalous ball and also whether it is heavier or lighter than each of the others.
There are no tricks or gimmicks. This is an honest problem. It can be solved. Below is my solution. I think (hope) my notation is self-explanatory.
Solution: (Click to enlarge photos. If it is still too small use the magnifying glass icon in the upper right on the referred page.)
Subscribe
