What do you want to bet that 6 months from now we will discovered that 75% of newly-insured exchange enrollees dropped out after paying premiums for a couple months.
My struggle with the McKinsey survey, which is a valuable resource, is that I don’t quite know what to make of the term “uninsured”.
People are defined as “uninsured” in the survey if they answer that their “primary insurance” in 2013 was uninsured. So, they could have been uninsured from January through July 2013, then got covered by an employer plan in August 2013, and still reported themselves as having been uninsured.
Alternatively, there could be (likely to be) recency bias. That is, I had insurance until October 2013, lost it, and then reported in April 2014 that I was mostly uninsured in 2013.
A lot of surveys of the uninsured have this limit.
Read the report. The authors do not estimate that 25% of exchange enrollees were previously uninsured. That would be impossible, for two reasons:
1) No distinction in the survey was made between on-exchange and off-exchange QHP’s.
2) They did not draw a random sample from the population of enrollees in the individual market, and then classify them as previously insured or uninsured. Rather, they drew separate, roughly equal sized, samples from the the previously insured and previously insured QHP eligible populations, and asked them about their current health insurance status.
Obviously, no statistical inference can be drawn from a survey designed this way, and the authors do not claim otherwise anywhere in the report. Read it.
The clipping has an inaccurate headline, but the link appears to be stale anyway. The actual survey is at: http://tinyurl.com/p969sxl.
P.2, first bullet point: “Of all
respondents who reported having selected a new ACA plan at the time of the April survey
(either on or off the exchanges), 26 percent reported being previously uninsured.”
So, McKinsey is counting all plans that have been newly offered in 2014, and compliant with ACA regulations, whether offered on exchange or individual market. (P. 1, fn 1 further clarifies this.)
So, McKinsey doesn’t really help us understand the flow of tax credits. I guess that McKinsey doesn’t really care about this because their clients are firms with high-salary employees and McKinsey wants to advise them how to respond to the ACA.
However, for the purposes of understanding what is happening to the uninsured, it doesn’t really matter whether they go on exchange or off exchange. For the purpose of public policy, we want to know how well the ACA is cutting down the number of uninsured.
P.2, fn 2, describes McKinsey’s definition of “uninsured” versus other surveys’ definition. I appreciate what they’ve done, but I still think there is a difference (for the purpose of estimating the effect of the ACA) between someone who was uninsured for seven months in 2013, January through July, versus someone who was uninsured June through December 2013 – also seven months.
Both will be reported as primarily uninsured by McKinsey, but the latter is more likely taking advantage of the guaranteed issue rules and suffers from a pre-existing condition.
It’s true that “we want to know how well the ACA is cutting down the number of uninsured”. Where on earth in the McKinsey survey do you see a statistic that would help us estimate this number? In fact I defy you to find a single statistic of any kind in the McKinsey survey. As the authors said in every version of their survey, their results were “directional” only.
Gallup, the other hand, is directly on-point. They ask a simple question: are you insured right now? Based on that question, they estimate that the number of uninsured adults has declined by over 11 million since September. They also estimate that half of all newly insured adults acquired their insurance through the exchanges. These estimates may or may not be accurate. But they at least they are actually statistics – with a defined statistical population, random samples and measurements.
Actually, the Gallup-Healthways survey is even more limited because it doesn’t tell us how long the respondents have been uninsured.
Especially when you see that the uninsured in Gallup-Healthways jumped up in the last quarter of 2013 (with no good explanation in the labor market), one is tempted to conclude that people were cancelled or dropped insurance in anticipation of Obamacare exchange enrollment. That is, a crowding-out effect.
That number is still heartwarming. These are millions who had no access to insurance beforehand, however it may outweigh the amount who had insurance and no longer do.
“Doctor salaries are a pretty significant part of the reason why the United States spends more per person on health care than any other developed country. About 56 percent of all medical spending goes to paying health care workers, rather than buying prescriptions, for example, or medical equipment.”
“Americans don’t actually go to the doctor a lot more than people in other countries. But when we do, our medical care costs more. Specific services, like MRIs and knee replacements, have significantly higher price tags when delivered in the United States than elsewhere.”
This would only be good if our healthcare is the best. Which it is not, and if you try to decrease their pay they cry fowl and say that lesser people would be come doctors. Yeah, ok. Let’s see how that turns out
Workers would rather save up for retirement than pay for health care.
That could be a matter of perception. Or it could be that workers don’t value medical care as much as they value retirement assets. Money you pay into retirement you expect to get back with interest. Money you pay for health insurance pays for benefits you hope you don’t need to access.
Was listening to Econtalk the other day with Scott Atlas and he said that those 15 million eligible, but uninsured didn’t realize medicaid was insurance for around 10 million of them.
“Survey: 25% of exchange enrollees previously uninsured.”
Big dent in the uninsured huh?
What do you want to bet that 6 months from now we will discovered that 75% of newly-insured exchange enrollees dropped out after paying premiums for a couple months.
Bet you’re right.
My struggle with the McKinsey survey, which is a valuable resource, is that I don’t quite know what to make of the term “uninsured”.
People are defined as “uninsured” in the survey if they answer that their “primary insurance” in 2013 was uninsured. So, they could have been uninsured from January through July 2013, then got covered by an employer plan in August 2013, and still reported themselves as having been uninsured.
Alternatively, there could be (likely to be) recency bias. That is, I had insurance until October 2013, lost it, and then reported in April 2014 that I was mostly uninsured in 2013.
A lot of surveys of the uninsured have this limit.
Read the report. The authors do not estimate that 25% of exchange enrollees were previously uninsured. That would be impossible, for two reasons:
1) No distinction in the survey was made between on-exchange and off-exchange QHP’s.
2) They did not draw a random sample from the population of enrollees in the individual market, and then classify them as previously insured or uninsured. Rather, they drew separate, roughly equal sized, samples from the the previously insured and previously insured QHP eligible populations, and asked them about their current health insurance status.
Obviously, no statistical inference can be drawn from a survey designed this way, and the authors do not claim otherwise anywhere in the report. Read it.
“…previously insured and previously uninsured populations QHP eligible populations…”
The clipping has an inaccurate headline, but the link appears to be stale anyway. The actual survey is at: http://tinyurl.com/p969sxl.
P.2, first bullet point: “Of all
respondents who reported having selected a new ACA plan at the time of the April survey
(either on or off the exchanges), 26 percent reported being previously uninsured.”
So, McKinsey is counting all plans that have been newly offered in 2014, and compliant with ACA regulations, whether offered on exchange or individual market. (P. 1, fn 1 further clarifies this.)
So, McKinsey doesn’t really help us understand the flow of tax credits. I guess that McKinsey doesn’t really care about this because their clients are firms with high-salary employees and McKinsey wants to advise them how to respond to the ACA.
However, for the purposes of understanding what is happening to the uninsured, it doesn’t really matter whether they go on exchange or off exchange. For the purpose of public policy, we want to know how well the ACA is cutting down the number of uninsured.
P.2, fn 2, describes McKinsey’s definition of “uninsured” versus other surveys’ definition. I appreciate what they’ve done, but I still think there is a difference (for the purpose of estimating the effect of the ACA) between someone who was uninsured for seven months in 2013, January through July, versus someone who was uninsured June through December 2013 – also seven months.
Both will be reported as primarily uninsured by McKinsey, but the latter is more likely taking advantage of the guaranteed issue rules and suffers from a pre-existing condition.
It makes a kind of thumbnail sense: 8 million new signups being touted, versus 6 million cancellation notices.
It’s true that “we want to know how well the ACA is cutting down the number of uninsured”. Where on earth in the McKinsey survey do you see a statistic that would help us estimate this number? In fact I defy you to find a single statistic of any kind in the McKinsey survey. As the authors said in every version of their survey, their results were “directional” only.
Gallup, the other hand, is directly on-point. They ask a simple question: are you insured right now? Based on that question, they estimate that the number of uninsured adults has declined by over 11 million since September. They also estimate that half of all newly insured adults acquired their insurance through the exchanges. These estimates may or may not be accurate. But they at least they are actually statistics – with a defined statistical population, random samples and measurements.
http://www.gallup.com/poll/168821/uninsured-rate-drops.aspx
http://www.gallup.com/poll/168548/newly-insured-2014-represent-adults.aspx
Actually, the Gallup-Healthways survey is even more limited because it doesn’t tell us how long the respondents have been uninsured.
Especially when you see that the uninsured in Gallup-Healthways jumped up in the last quarter of 2013 (with no good explanation in the labor market), one is tempted to conclude that people were cancelled or dropped insurance in anticipation of Obamacare exchange enrollment. That is, a crowding-out effect.
Reducing smoking, drinking alcohol, salt… would save 37 million lives over 15 years.
Would people actually live longer? Or would it just feel longer when they’re deprived of things that make life worth living?
love me some sigrits.
Yeah, but who wants to reduce all those fun filled activities
That number is still heartwarming. These are millions who had no access to insurance beforehand, however it may outweigh the amount who had insurance and no longer do.
One must not outweigh the other.
“Doctor salaries are a pretty significant part of the reason why the United States spends more per person on health care than any other developed country. About 56 percent of all medical spending goes to paying health care workers, rather than buying prescriptions, for example, or medical equipment.”
Exactly.
“Americans don’t actually go to the doctor a lot more than people in other countries. But when we do, our medical care costs more. Specific services, like MRIs and knee replacements, have significantly higher price tags when delivered in the United States than elsewhere.”
This would only be good if our healthcare is the best. Which it is not, and if you try to decrease their pay they cry fowl and say that lesser people would be come doctors. Yeah, ok. Let’s see how that turns out
The current doctors would certainly change their spending habits.
Re: Doctor’s pay:
http://www.bestmedicaldegrees.com/salary-of-doctors/
Shorten their education then. You could remove 2 years of undergrad and save 50k over that course.
That would be pennies over the course of hours worked.
Workers would rather save up for retirement than pay for health care.
That could be a matter of perception. Or it could be that workers don’t value medical care as much as they value retirement assets. Money you pay into retirement you expect to get back with interest. Money you pay for health insurance pays for benefits you hope you don’t need to access.
Interest is seen as the incentive to retirement.
Was listening to Econtalk the other day with Scott Atlas and he said that those 15 million eligible, but uninsured didn’t realize medicaid was insurance for around 10 million of them.