[Editor: Update from the Daily Mail and our analysis on these numbers]
[Update: Why these numbers matter]
One of the main issues of ACA (Obamacare) is that people do not know how many new people have enrolled on the federal exchange program. I have attempted solve that issue by making projections on how many people enrolled on the federal exchanges based on state exchange data. I have provided the data on how I made these projections below.
Measuring the number of people that have signed up for an actual policy under Obamacare has been a difficult thing to do. The White House and HHS Secretary Kathleen Sebelius claim not to know the number. WonkBlog noted that the states and the federal government are measuring “enrolled” differently. I’ve managed to cut through the difficulty that WonkBlog notes. Finally, the Obama administration has promised to give the actual numbers out on a monthly basis.
Aaron Strauss provided a table to numbers from the states on a Google Doc file that anyone can view. I have updated those numbers and revised the table to reflect the my knowledge known as of October 8, 2013 at 9 pm. I have provided links for every number I pulled, attempting to ensure the data is as transparent as possible. I have also added conversion categories that I will explain below. I will make revisions to those numbers after the 2nd week of ACA has passed so I can include 2 weeks of data for comparison. Hopefully, the state exchanges will continue to provide weekly updates on their progress so a proper extrapolation can be made for the federal exchange. Unless otherwise linked, all numbers I refer to are from my tables linked above.
State Exchange Numbers:
Projections are that the White House received 9 million unique visitors on the Federal Insurance Exchange: Healthcare.gov. A unique visitor is the person who visits the website; their visit counts only once. The total clicks on a website is how many link clicks a unique visitor makes. Unique visitors is the most useful metric since it gives us an idea for exactly how many IP addresses visited the exchange. Other estimates have been closer to 8.6 million unique visitors. My estimates of the state exchanges, directly using numbers states have provided, is that the state exchanges received just over 6 million unique visitors on the various state exchange websites (it’s obviously higher, but the official data only gets us this far for the time being).
The 17 state run insurance exchanges have reported roughly 100,000 people fully enrolled in ACA (Obamacare). This figure means that during the first week 100,000 people are either fully enrolled with health insurance or they have filed a complete application with full or pending payment. Incomplete applications are not included into this category.
The 17 state run insurance exchanges have also reported a total of just under 234,000 new accounts set up. An account is where a unique visitor to the state exchange website set up a name and password on the exchange website. This number will vary widely from state to state. The variance is due to differences in the state exchanges on setting up an account. Some states require you to set up an account before viewing rates and plans, others allow you to simply browse and only set up an account once you are ready to purchase.
The conversion percentages are my analysis of the numbers provided by the states. First, there is the conversion percentage from a unique visitor to an enrolled customer in the exchange. This number measures how many unique visitors are turned into an enrolled customer on the exchange website (the main goal of the ACA). Second, there is the conversion rate of unique visitors into registered accounts. This number measures how well the website is turning unique visitors into account holders on the website with a name and password. Finally, there is the conversion rate of account holders into enrolled insurance customers. This number measures the website’s ability to turn an account into an enrolled paying customer.
Overview of the State Exchange Numbers
An overview of the numbers from the first week of ACA (Obamacare) in action reveals:
- State exchange websites received approximately: 6,000,000 Unique Visitors (UV’s)
- State exchange websites had approximately: 100,000 enrolled paying/pending payment customers
- State exchange websites had approximately: 234,000 new accounts created on them
- Average conversion rate for turning UV’s into enrolled customers: 1.68%
- Average conversion rate for turning UV’s into accounts: 3.89%
- Average conversion rate for turning accounts into enrolled customers: 43%
Overall, the conversion rate for turning website views into an account holder is low. Account registration is usually the first level of interaction a user has with an exchange site (aside from those states that allow window shopping prior to registering an account). These numbers tell us that 96.1% of UV’s are remaining UV’s and not forming an account or enrolling in any care. 98.3% of UV’s and account holders are not enrolling into any plan, even though they may have an account. To be fair, low numbers should be expected this early in the program. If these low numbers continue into the coming months, red flags should be raised.
As I mentioned above, this is the product of the 17 state exchanges minus Utah and Idaho. Utah is running the exchange for businesses, but not for individuals. Idaho is slated to be a state run exchange in the fall of 2014. Until then, Idaho is being run by the Feds. Idaho did provide some numbers on UV’s. As my spreadsheet shows, there are many gaps in the data that need to be filled, so the projections I will make on the federal exchange should only be considered a rough draft of what to expect from the Healthcare.gov site. The more public information the state exchanges release, the better idea we can get of what to expect from the federal exchange. Similarly, we need further data on the state exchanges to determine how effective they are at converting UV’s into enrolled members and account holders.
For now, let’s look at the 5 best and worst state exchanges:
The 5 Best State Exchanges at getting a person enrolled were (based on conversion rates):
- Minnesota: Conversion rate: 22%. I actually anticipate Minnesota to come down from this high figure after better figures are received from the state. 6,600 were reported to be enrolled in Minnesota. Minnesota reported many glitches and the UV number of 30,000 was low for a large state.
- Kentucky: Conversion rate: 9.89%. Kentucky probably had the single best week of any state in the US. The Governor reported the official numbers that they received. The report was 17,314 people enrolled. Officials managed to convert nearly 10% of UV’s that visited the exchange site into paying customers. 14.4% of UV’s made accounts on the exchange site. While this number may seem a tad low, Kentucky managed to turn a whopping 71.1% of those who made an account into an enrolled customer. Kentucky allowed people to window shop without registering an account, which explains the lower account percentages. Kentucky has been a model of success for exchange websites so far.
- Hawaii: Conversion rate: 6.07%. Hawaii is only reporting the first two days of their exchange being open. In those two days Hawaii managed to turn 1,181 of its nearly 20,000 UV’s into enrolled members [Ed. note: We have just learned this morning that Hawaii was misreporting its numbers. They did not enroll a single person. This moves them into a tie for 1st with Oregon for the worst health exchange]
- Washington (state): Conversion rate: 5.71%. Washington had a solid day. Managing to turn nearly 9,500 of its 165,000 UV’s into enrolled members.
- Rhode Island: Conversion rate: 2.45%. Rhode Island reported its 1st and 3rd days of enrollment. They managed to get 794 people enrolled.
The 5 Worst State Exchanges at getting a person enrolled (based on conversion rates):
- Oregon: Conversion rate: 0%. Oregon couldn’t manage a single enrollment figure over the course of the week. This is despite having 234,000 UV’s visit the site. Oregon had one of the “glitchiest” sites in the nation and people had trouble with nearly every feature on the site. @CoverOregon promises that the site will function better the second week. Oregon gets the top spot because there are no official counts from the first week that they registered an enrollment OR an account.
- Vermont: Conversion rate: 0%. Vermont is a nudge better than Oregon because they managed to get accounts set up. However, the people in charge of the state exchange said they hadn’t seen one enrollment, leaving the rate at zero.
- Massachusetts: Conversion rate: 0.07%. Massachusetts only managed to get 1,134 enrolled after receiving 1,600,000 UV’s on their exchange website. Part of the explanation here is that Mass. already had a low uninsured population. Only 0.43% of the UV’s even bothered to set up an account.
- Maryland: Conversion rate: 0.19%. Maryland simply performed horrible across the board. They were only able to get 326 people enrolled with a UV number of 170,000. They performed better at getting accounts, with an account conversion rate of 7.96%. However, Maryland was only able to get an abysmal 2.4% of its accounts to turn into an enrollment.
- Connecticut: Conversion rate: 1.56%. Conn. fell just under the average conversion rate of 1.68% across all the state exchanges. They were able to get 1,157 enrollments off of 100,000 UV’s.
Federal Exchange Enrollment Projections
The Federal Exchange remains mum on how many people it signed up in 27+ states over the first week. However, it is possible to extrapolate from the state exchange data to get an idea of where the Feds could potentially stand after the first week of ACA enrollment. I have four ways to project how the Feds could or should be doing in enrollment and accounts: 1) Overall comparison, 2) Best case scenario, 3) Worst case scenario, and 4) Best data comparison.
1) Overall Comparison Projection
To achieve this number, you take the overall numbers from the state analysis above and project the average across federal exchange. For reference, the overall conversion numbers above were: 1.68% for enrollment and 3.89% for accounts. The UV number for the federal exchange is 9 million. With those presumptions set:
- A enrollment conversion rate of 1.68% across 9 million UV’s would mean: 148,500 newly enrolled people
- A new accounts conversion rate of 3.89% across 9 million UV’s would mean: 350,100 new accounts set up
- If the Feds had roughly those two numbers, that would give them an accounts to enrollment conversion rate of 42.4%.
These figures would likely be an average job by the federal exchange in the first week. This scenario assumes that the federal exchange, dealing with people across 27 states, would average the same as the other 17 state exchanges.
2) Best Case Scenario Projection
The best case scenario projection takes the five best state exchanges and assumes that the federal exchange would perform above the average overall comparison projection above and match what the best states would did during the first week. The top five states enrolled 35,341 people with 422,160 UV’s. The top five states averaged a conversion rate of 8.37% for enrollment. Accounts information is unavailable for the full top five states on conversion. 8.37% projected across 9 million would mean the federal exchange enrolled potentially 753,000 people in the first week of ACA.
This rosy picture would be the top end of what the federal government could be expected to produce in a single week. Given the glitches and problems with the federal exchange website, this number seems unlikely.
3) Worst Case Scenario Projection
The worst case scenario projection works just as the same as the above, except the average is the five worst performing state exchanges, in terms of conversion rates. The five worst states enrolled 2,617 people. They also have 2,146,500 UV’s. This would give the five worst states an average conversion rate of 0.12%. Accounts information is unavailable for the five worst states as well. 0.12% projected across 9 million would mean the federal exchange enrolled 11,000 people in the first week of ACA.
If the federal exchange only manages to enroll 11,000 people in the first week, with 9 million UV’s, red flags need to go up immediately. 11,000 spread across 27 states would be a disgraceful number. It would be hard to imagine the federal exchange hitting this low of a number in the first week out with the state-run exchanges obtaining 100,000 by themselves (even if California, New York, and Kentucky account most most of the state exchange numbers).
4) Best Data Projection
The best data projection, admittedly, cherry picks certain states out of the state exchanges to provide a comparison. Although Vermont and Oregon had complete data for their exchanges, I have cut them here to avoid a 0% slant on the conversion rate. The states with the best and most complete data are: Kentucky, California, Maryland, Massachusetts, Rhode Island, and Washington. New York could be included, but the UV numbers given by that state are only a two hour total for the first day. New York saw two million in that two hour period making it impossible to extrapolate what a week total would be from that state. Raw numbers for the other six states:
- These six states had 132,241 accounts registered on the exchanges, they enrolled 45,331 into insurance polices, and had 3,130,143 UV’s on their websites.
- The six states have an enrollment conversion rate of 1.45%, and account registration conversion rate of 4.23%, and an account to enrollment conversion rate of 34.3%.
- An enrollment conversion rate of 1.45% spread over 9 million UV’s produces: 130,500 people enrolled
- An account registration conversion rate of 4.23% spread over 9 million UV’s produces: 380,700 accounts
Compared to the overall comparison projection, the complete data projection shows slightly lower numbers for the number of people enrolled and a slightly higher number of account registrations. There is a 9 point drop on the account to enrollment conversion rate, from 43% to 34%. The enrollment conversion rate falls as well, from 1.68% to 1.45%.
First, it is highly likely that the Obama administration is refusing to release weekly numbers for Obamacare because the conversion rates for UV to enrollment are low, potentially in the 1.4% – 1.7% range, meaning 98% of website visitors are not signing up for the program. While, in the long term, this is meaningless because conversion rates would be expected to be low this early on, the political criticism on 1% conversion rates would be tough. The fallout could potentially give the public even more negative reactions to Obamacare. These numbers could also harm the administration’s position that Obamacare does not need to be delayed until it works properly. The real worries come if low 1% conversion rates continue in the coming months. Low conversion could be compounded by the exchanges providing bad and faulty information to insurance providers.
Second, while there may be an initial rush in enrollment numbers the first few weeks, I would anticipate those numbers to level off. There are two other potential time periods where a rush would be more dangerous to the websites than right now. First, from Dec. 1-15th, and second, in the last two weeks of March 2014. Many of the exchanges noted that if a person wanted to receive coverage that started on January 1st, they should sign up by December 15th. Navigators will be sure to push this point. If people wait through the holidays, look for March to be the big sign up month. Open enrollment closes on March 31st. I would anticipate the last of March being a rush for healthcare enrollment. During the other periods of time, there should just be a normal level of insurance purchases.
Finally, it will be interesting to see the second week’s data come trickle in. If the state exchanges experience bad times again, it will hurt enrollment in many states where conversion rates were already low. Navigator training is still taking place in many of these states, which only adds to the ill-prepared nature of Obamacare implementation. There is plenty of ground to cover in the coming weeks.