Gaining
Coverage in a Conservative World: A Consideration of the Results of the
Affordable Care Act
In 2014, the
Affordable Care Act (ACA) went into effect, and with it came the option for
states to expand their Medicaid coverage. The expansion was originally intended
to be mandatory, but the Supreme Court declared that unconstitutional, so
states were given the option to expand Medicaid coverage to include people
whose incomes are 138% of the Federal Poverty Line or below. Even if the states
did not choose to expand in 2014, they are still able to expand it now, using a
Section 1115 waiver, which will provide the same funding guarantees that the
original expansion did (Rudowitz and Musumeci,
2015). The ACA promised 100% funding for all expansions of Medicaid up
to the 138% level between 2014 and 2017, and then funding decreases to 95% for
the 2017 year, 94% in 2018, 93% in 2019, and 90% in 2020. Those states who took
the waivers receive transition funding until they reach the 138% level, and
then their three years of full funding will start (MACPAC, 2017).
This paper will
cover three schools of existing literature; those against the ACA, those who
fundamentally agree with the ACA, but disagree with the methods, and those who
feel that the ACA is adequate as it is written. Next, the hypothesis is presented,
and the variables are operationalized. Finally, the sources of data are
identified, and the data analysis is planned.
The
ACA has been a hot topic of debate since its inception, but two of the most
fiercely debated aspects of the act are its effectiveness in reducing costs for
the average Medicaid recipient and its fiscal impact on the states. Under the
ACA, states had the option to expand Medicaid up to 138% of the Federal Poverty
Line (FPL) with federal funding. However, after three years, federal funding
for the expansion slowed, and in 2020, states will have to take on 10% of the
cost of the expanded Medicare. Ideally, the ACA would have led to lower costs
for the states because they would have fewer uninsured citizens, and it would
have lowered costs for patients because more people were paying into the
Medicaid system. So how have varying levels of Medicare expansion under the
Affordable Care Act actually affected state expenditures and patient costs in
the United States?
Scholars argue
whether or not the ACA has succeeded in lowering some of the costs of Medicaid,
forming three schools of thought: 1. The ACA is inherently flawed, and there is
no correcting it, 2. The ACA is flawed, but it provides coverage that the
market cannot, and 3. The ACA is a good piece of legislation and just needs
some time to smooth over the kinks from its rollout. Each of these schools has
a different answer to the research question, and each is tied to their overall
opinion of the ACA.
The
first school believes overwhelmingly that the ACA was a failure, and many of
the scholars support the idea that it should never have been proposed. It was a
polarizing issue from its introduction on the floor of Congress, and continues
to be as Republicans aim to repeal it. The scholars within this field disagree
about what exactly the problem is with the ACA, but they all support the
dismantling of the ACA as it currently stands. They theorize that the best
outcomes in healthcare are created through methods other than expanding Medicaid.
Gaffney and Waitzkin (2016) believe that our current healthcare system as a
whole is flawed, starting from its inception. They argue our rising healthcare
costs can be attributed to the powerful lobbying groups in the United States,
as well as our flawed political system. Gaffney and Waitzkin criticize the
consumerist nature of the American health system, noting that it drives up
prices and prevents the single-payer system from emerging. Gaffney and Waitzkin
reiterate the merits of a single-payer system, from a more simplistic
bureaucracy, to increased savings for the government, to better coverage for
the consumer. Sommers et al. (2014) also criticize the consumerist system, but
they address the failure of the ACA to reach those that need it most as well.
The expansion of Medicaid from the ACA created a class of Americans who can
sporadically afford healthcare, and therefore may switch back and forth between
private insurance and Medicaid several times as their circumstances change.
Despite the ACA’s attempt to lure more doctors into accepting Medicaid by
raising the reimbursement rate, Sommers et al. predict that there would only be
a small increase in doctors who accept Medicaid. They admit that the ACA is
noble in its cause, but recognize that many states are still skeptical about
the fiscal feasibility and that the federal government will provide the
benefits at the levels that it claims. Thompson (2012) agrees with Gaffney and
Waitzkin that the whole system is flawed, but he takes a greater stance against
the idea of cooperative federalism. Thompson believes that the Obama
administration’s plan to give the states greater control in the process doomed
the ACA before it even hit the president’s desk. Thompson finds the coercion of
the states to expand Medicaid problematic, and he faults the federal government
for not declaring a plan for when a state opts out of Medicaid entirely. The
crux of his argument becomes a non-problem with the decision of the Supreme
Court in National Federation of
Independent Businesses v. Sebelius, but that does not detract from
Thompson’s initial concerns. Much like Gaffney and Waitzkin, Thompson also
recognizes the merits of a single-payer system, but he does not push the issue
as far.
The
second school is compelled by the progress that the ACA has made, but finds
flaws in the details of the law. These scholars claim that the outcomes could
have been better with smoother implementation, or with widespread adoption of
Medicaid expansion, but they generally agree that the law was created in good
faith. Persad (2015) argues that while the law does good and promotes a
positive worldview, its priority setting is flawed. No person’s life is allowed
an actuarial value higher than another’s, which can lead to higher overall
costs as life-extending procedures are prioritized over quality of life
increases. Persad recognizes the potential problems associated with assigning
one life a value over another, but believes that the savings are worth the
moral cost. Kreuter et al. (2014) found the most lacking feature of the ACA is
that there is no way to contact those who were eligible to enroll in Medicaid
or get a plan through the marketplace. Shartzer, Long, and Anderson (2016)
disagree, instead focusing on the gaps that the ACA leaves behind. They note
that the Medicaid expansion will lead to less funding for programs for the
uninsured, leaving them hanging by a thread. Shartzer, Long, and Anderson do
not find fault in the actuarial values, as Persad does, like Gaffney and
Watzkin, do acknowledge the effects of the consumerist system, noting that gaps
in affordability due to the market are causing many of the coverage holes.
The
third school finds that the ACA was a valiant effort, and has produced the
results that can be expected from a new piece of legislation. They believe that
the outcomes are the best that one could hope for from a piece of legislation
this expansive and contentious. Kowalski (2014) argues that even if prices were
higher in the healthcare exchanges, the impact of the ACA would still be
encouraging. She acknowledges the shaky and staggered rollout of the bill, and
attributes the bill’s flaws to the website problems and poor press that it
received. Hill (2015) focuses more on the Medicaid portion of the law, claiming
that individuals under Medicaid would spend half as much on healthcare costs as
compared to those enrolling in the marketplace. Knopf (2012) identifies reasons
why states would potentially turn down the 100 percent funding that the ACA
would provide to states, and ultimately concludes that no sane state would
reject the Medicaid expansion, as it has the potential to balance state
budgets. Frean, Gruber, and Sommers (2016) agree with Hill and Kowalski’s
sentiments, lauding the benefits of the Medicaid expansion, and noting the
specific increase in enrollment of previously eligible Americans. Much like
Kreuter et al., they note that the previously-qualified individuals showed the
greatest capacity for growth. Barrilleaux and Rainey (2014) corroborate Knopf,
arguing that the strongest and only real reason for the opposition of the ACA
is partisanship. They note that many governors cite fiscal concerns, much like
Sommers et al. does, but Barrilleaux and Rainey provide evidence that the
states are much more likely to profit from the venture. Blumberg and Holahan
(2016) concur, showing predicted spending data for states who failed to expand
Medicaid. Blumberg and Holahan also agree with Shartzer, Long, and Anderson
that the coverage gaps are large, but Blumberg and Holahan believe that the ACA
is adequate for now.
These
three schools, when looked at together, paint a complicated picture of the ACA.
It is a flawed document, and its rollout was shaky at best, but overall it is
doing a great service to the poor in America. Perhaps there will be better
bills in the future, but this is the one that we have now, and this is how it
was passed by the United States Congress, so this is how it is going to stay.
What these studies and articles have missed, is to what degree has each state
succeeded in improving the lives of its residents. Many of them were also
published just after the bill went into effect, in mid-to late 2012, so there
is a dearth of recent knowledge as well.
I
plan to fill this knowledge gap by focusing on six states’ implementation of
the Medicaid expansion; Montana, Arizona, Nevada, Colorado, New Mexico, and
North Dakota. These states have fairly similar cultural values, racial makeups,
and climates, and they all expanded Medicaid. The states will be compared using
a most similar systems approach in order to best utilize the homogeneity of the
states.
Hypothesis
I hypothesize that all states will show greater
federal funding, most states will report a plurality of expanded Medicaid and
non-Medicaid insurance coverage, but also increased costs from the states At
least one state will have reduced insurance coverage and increased costs for
the state, and at least one will show no significant change in the observed
variables.
Operationalization
When discussing any
research, it is important that one is accurate and precise in their wording,
and one must use indicators that measure what is being researched. Precision is
key because it allows for the research to be generalizable on the same level of
analysis, and accuracy is important because it allows for the experiment to be
repeated by other researchers. If the indicators do not measure the variables
that the researcher wants to measure, the findings will not be valid. This
researcher strives to achieve valid and reliable results, and will therefore
detail the variables, indicators, and data sources further into this paper.
Medicaid is
defined as a health insurance alternative for children, pregnant women, low
income families, those with disabilities, and those with incomes at or below
138% of the Federal Poverty Line (FPL). The Affordable Care Act (ACA) is
defined as the Patient Protection and Affordable Care Act, a bill passed in
2010 by the Obama administration, colloquially called “Obamacare.” The FPL is
defined as the level at which the government determines eligibility for
programs such as Medicaid. The current FPL for one individual is $12,060. The
states are defined as the six specific states being studied, namely Arizona,
Colorado, Montana, Nevada, New Mexico, and North Dakota. Insurance is defined
as health insurance provided by an accredited insurance company, or a government
plan such as Medicaid. Implementation is defined as the point at which the
state can be reasonably assumed to offer coverage to 138% of the FPL.
The indicators discussed in the next paragraph
represent the independent variables in this case, affordability and reach, and
the dependent variable is the more abstract concept of success. If a plurality
of the states show the same or lower costs for the state governments and
greater coverage in the more recent data set, then the ACA will be deemed a
“success” by the researcher. If a plurality of the states saw no or negligible
effect on costs and coverage, then the ACA will be shown to have no effect on the
observed variables. If a plurality of the states has higher costs and less
coverage in the more recent data set, then the ACA will be shown to not be a
success, and has failed its goal. If the variables show mixed results in a
plurality of states, then the ACA will be shown to need further study, but
shall be preliminarily deemed a failure, pending further investigation.
This research will
be completed by using several different indicators; the number of people at or
under 138% of the FPL insured by Medicaid before and after implementation by
the states, the percentage of total people insured before and after
implementation, and each state’s Medicaid spending per capita before and after
implementation, and the amount of federal funding received by the state per
capita before and after implementation. These indicators help measure the
variables of affordability and the reach of the bill. Both need to be measured,
as it is very easy for Medicaid to be more affordable to states and the federal
government if they decrease its reach. The
indicators were drawn from and inspired by the work of Kowalski (2014),
and federal funding was included as an indicator because of the work of Knopf
(2012). The Politics of Need
(Barrilleaux & Rainey, 2014) was consulted to find more variables, but they
had been covered in Kowalski’s methodology, and seemed redundant to add.
General data about the federal and state shares were found on the Medicaid and CHIP Payment and Access Commission
website (2017), and the more specific indicators were formed based on that
information. The first indicator- the number of people enrolled in
Medicaid that fall under 138% of the FPL- will be obtained from a combination
of enrollment data from the Kaiser Family Foundation (2017) and census data (Population Division of the U.S. Census Bureau,
2016). The “before” data will be taken from 2013, and the “after” data will be
taken from either one and a half years after the expansion began, or January of
2017, whichever occurs earlier. This is to ensure the fairness of the
indicator, while still allowing for the research to be completed promptly. The
dates were also chosen based on the findings of Vistnes and Cohen (2016), as
their data did not go far enough to determine the real outcome of the ACA. The second
indicator- the percentage of the population that is generally insured- will use
the same data, but will look at it more broadly, comparing the number of people
insured by any group; Medicaid or private insurance. The data will be sampled
from 2013 and one and a half years after implementation, or January 2017,
whichever comes first. The third indicator- each state’s Medicaid spending-
will be created using spending data from the Kaiser Family Foundation and
census data. Once again, the data will be taken from 2013 and either one and a
half years after implementation, or January 2017, whichever is earlier. The
fourth and final indicator- the amount of federal funding received- will be
created by combining information gathered from Congressional Research Service
documents (Mitchell, 2014), and census data.
These variables are good indicators of the success of
the ACA because they are the stated goals of the bill. Ideally, the bill would
make healthcare cheaper for states and the federal government, and it would
provide better coverage to more people. The number of people insured by Medicaid
at or below 138% of the FPL will show the bill’s intent regarding Medicaid. If
this number does not increase, then the bill did not help the poorest Americans
become insured. The number of people insured regardless of socioeconomic status
shows the primary goal of the bill- making sure every American is
insured. State spending shows whether the ACA succeeded in making universal
healthcare a viable goal for states to pursue. Federal funding for each state
shows whether the federal government is taking its fair share of the costs. If
the ACA works as planned, then the federal spending would increase
dramatically, and the state spending would stay the same.
Data for these indicators are easily obtainable and
offer clear measurements of the success of the bill, and are therefore likely
to be valid. They are numerical, interval figures, and are easily comparable in
order to observe change over time. The data used in this research will be
aggregate data, as sampling such a large population would exhaust the
researcher’s time and money before any analysis could be completed. The group
being studied is the demographic group of Medicaid recipients in a set of given
states (Arizona, Colorado, Montana, Nevada, New Mexico, and North Dakota). It would be helpful to include variables like
health of the population, but these are easily influenced by things outside of
the scope of the ACA, and are difficult to quantify into indicators. Even if
that data could be quantified, it would be incredibly difficult to obtain them
due to patient privacy concerns. Another variable that was considered was
affordability of Medicaid for the patient, but few states charge anything for
Medicaid, so including it would only unbalance the measurements.
The researcher hypothesizes that the states of Arizona,
Colorado, Montana, Nevada, New Mexico, and North Dakota will show mixed
results, but all will show greater federal funding. The states will likely report
a plurality of expanded coverage in both sectors, but also increased costs from
the states as they reallocate the federal funds to non-healthcare related
areas. At least one state will be judged to have failed, and at least one will
show no change in the observed variables.
There
are several distinct data that will be collected for this research; population
of the six states, number of people insured, number of people insured
specifically by Medicaid, federal funding provided to each state, and Medicaid
spending for each state. The data that will be collected in coordination with
the research will be aggregated from four main sources. These sources are the
Kaiser Family Foundation, the US Census, the Medicaid
and CHIP Payment and Access Commission, and the Congressional Research Service (Kaiser
Family Foundation, 2017; Mitchell, 2014; Population Division of the U.S. Census
Bureau, 2016; The Medicaid and CHIP Payment and Access Commission, 2017). Population
data will come from the census, and if that is not available for some reason,
then the data will be collected from the UN Demographic Yearbook. Federal
funding will be measured from the Medicaid Access Commission and the
Congressional Research Service, and if those sources fail, then the researcher
will extrapolate the data from federal and state budgets. State spending will
also be measured using data from the Medicaid Access Commission and the Kaiser
Family Foundation. State spending will not include the funds given to the states
by the federal government- it will only be the money that states raised and
spent on their own. The total number of insured Americans and the number of
people insured by Medicaid will be collected using the Kaiser Family Foundation’s
website. The data will be collected in January of 2013 and either January 2017
or one and a half years after the Medicaid expansion began, whichever is
earlier, in order to standardize the time frame without limiting it too much. The
time frame also only includes the Obama administration to avoid adding in the
extra variable of a new president. The existing literature’s main flaw is in
their short-term view, only looking at six months to a year of data before
making generalizations about the bill.
The
researcher did not anticipate any problems finding the data until late January,
when the Trump administration’s policies became apparent and data began
disappearing from the public view. The researcher now anticipates problems
finding data sourced from the federal government, such as census data and
Congressional Research Service documents. Private data aggregators may capture
the needed data and provide it for free, or they may place it behind a paywall,
which is a significant barrier for the researcher.
The data collected above will be
combined to measure the indicators. Each piece of data will be divided by the
state’s population to make the data per capita, or a percentage of the
population. The data will then be made into percent change by dividing the
later data by the newer data and subtracting one: . This modification of the raw
data will ensure the validity of the data, and the reputable sources from which
the data are gathered ensures the data’s reliability. Data for these indicators
are easily obtainable and offer clear measurements of the success of the bill,
and are therefore likely to be valid. They are numerical, interval figures, and
are easily comparable in order to observe change over time. The indicators to
be measured are percentage of the population insured by Medicaid, percentage of
the population insured by any group, federal funding per capita, and each
state’s Medicaid spending per capita. It would be helpful to include variables
like health of the population, but those are easily influenced by things
outside of the scope of the ACA, and are difficult to quantify into indicators.
Even if that data could be quantified, it would be incredibly difficult to
obtain them due to patient privacy concerns. Another variable that was
considered was affordability of Medicaid for the patient, but few states charge
anything for Medicaid, so including it would only unbalance the measurements.
All the data collected will be quantitative, and will
therefore not require much coding to prepare for analysis. The 2013 and 2017
data will be compared, and then placed into graphs to create a visual
representation. Percent increases and decreases will be calculated from the
2013 and 2017 data, and then it will be weighted. The larger the final number
for each state, the more success that state achieved. Per capita federal
funding will be weighted the least (relative to the desired impact), with a
multiplier of one: . Per capita state spending will be weighted with a multiplier of point
nine: . The divisor weight in this case reflects the expected outcome of
lowered state costs. If the weight were a multiplier, then it would
artificially inflate the final value, as the researcher has deemed an increase
in state spending to be an indicator of failure, not success. Rates of
insurance will be given a weight of one-point five, as that was a goal of the
bill, but not the most important one: . Rates of Medicaid enrollment will be given a weight of two, as that is
the focus of the portion of the bill being studied: . The final equation will look close to this: . The
state with the greatest degree of success will be the most successful, and the
state with the lowest degree of success will be the least successful, but
neither will necessarily be a success or failure. Failure would be determined
by the ,,
or values being negative at all, or the value being positive and greater than eight
percent. If the ,,
or values
are negative, then federal funding would have decreased despite an increase in
expected federal funds, or the overall percentage of insured persons would have
decreased, and the goal of the ACA was to increase the number of insured
persons by increasing federal funding to the states. If the value is greater than positive eight percent,
then state expenditures outpaced average spending increases when Medicaid
enrollment was not bolstered, even outpacing recession numbers. The average
spending growth in the 20 years before the ACA was less than seven percent, and
even during the 2008 recession, when enrollment jumped by 8 million in two
years, the annual spending growth was only six-point eight percent (The Medicaid and CHIP Payment and Access Commission,
2017). The researcher gave the states an extra one
percent of spending growth to account for any infrastructure and developmental
costs associated with greater enrollment, so any annual spending increases
greater than eight percent must be outliers. If a state has all positive ,,
and values and a negative value, then it will be a success.
The hypothesis was that at most four states will have
succeeded in expanding Medicaid and insurance overall, one state will
have failed, accumulating more costs to the state and not raising enrollment,
and one state will not show significant changes in any of the observed
variables. This can be tested by using the formula
, and
looking at the indicator data. Significant changes in the observed variables
will be a change of for insured populations, as fluctuations are
expected regardless of new legislation, and for federal funding and state spending. The
limitations with this research is that it cannot necessarily be generalized to
other states, but the research methods may be applied to other states for more
state-specific results. If all the states are shown to be “successes,” then it
may be extrapolated that the ACA was successful in the west and southwest, but
more research must be conducted before it may be applied nationwide. This
research design could be improved by adding more specific variables such as
quality of healthcare received and number of hospital visits per capita. It
could also be improved by looking at a longer time frame and sampling more than
twice. However, those changes would have exceeded the researcher’s time and
financial budget, and the data would be locked behind paywalls.
Conclusion
This research is not about gathering
data; it is about changing lives. The people who rode the line between Medicaid
and no insurance in the pre-ACA years are now able to get life-saving medical
care, without the bankruptcy-causing bills. However, the new 138% line has
created a new group of people who ride the line, albeit a smaller one. Although
an affirmation of the hypothesis is always nice, real results that provide a
path for states is what will help the most people. If the researcher is wrong,
and all states failed, then we will know that there is work to be done, and new
legislation must be drafted. This project will soon be upstaged by
better-funded, longer term studies, but for now, it is the most up-to-date,
accurate portrayal of the ACA. It acts as a stepping stone for new healthcare
research, and will give rise to new ideas based on its limitations and
shortcomings, just as this research was inspired by what was missing from
previous studies.
Barrilleaux, C., &
Rainey, C. (2014). The politics of need. State
Politics & Policy Quarterly, 14(4),
437-460. doi:10.1177/1532440014561644
Blumberg, L. J., &
Holahan, J. (2016). Early experience with the ACA: Coverage gains, pooling of
risk, and medicaid expansion. The
Journal of Law, Medicine & Ethics, 44(4),
538-545. doi:10.1177/1073110516684784
Frean, M., Gruber, J.,
& Sommers, B. D. (2016). Disentangling the ACA's coverage effects - lessons
for policymakers. New
England Journal of Medicine, 375(17),
1605-1608. doi:10.1056/NEJMp1609016
Gaffney, A., &
Waitzkin, H. (2016). The affordable
care act and the transformation of US health care/la ley de cuidado de la salud
asequible y la transformación del sistema de cuidado de la salud
estadounidense. Cuadernos
De Relaciones Laborales; Madrid, 34(2),
239-261. doi://dx.doi.org.ezproxy.neu.edu/10.5209/CRLA.53456
Hill, S. C. (2015).
Medicaid expansion in opt-out states would produce consumer savings and less
financial burden than exchange coverage. Health
Affairs, 34(2), 340-349.
doi:10.1377/hlthaff.2014.1058
Jost, T. S. (2016).
Health care reform and A failed vision of bipartisanship. Health Affairs; Chevy Chase, 35(10), 1748-1752.
doi://dx.doi.org.ezproxy.neu.edu/10.1377/hlthaff.2016.1080
Kreuter, M. W., McBride,
T.,D., Caburnay, C. A., Poor, T., Thompson, V. S. L., Alcaraz, K. I., . . .
Casey, C. (2014). What can health communication science offer for ACA
implementation? five evidence-informed strategies for expanding medicaid
enrollment. Milbank
Quarterly, 92(1), 40-62.
doi:10.1111/1468-0009.12040
Mitchell, A. (2014). Medicaid financing and
expenditures ().Congressional
Research Service.
Persad, G. (2015).
Priority setting, cost-effectiveness, and the affordable care act. American Journal of Law
& Medicine, 41(1),
119-166. doi:10.1177/0098858815591511
Population Division of
the U.S. Census Bureau. (2016). Annual
estimates of the resident population for the united states, regions, states,
and puerto rico. ().U.S.
Department of Commerce.
Shartzer, A., Long, S.
K., & Anderson, N. (2016). Access to care and affordability have improved following
affordable care act implementation; problems remain. Health Affairs; Chevy Chase, 35(1), 168N.
doi://dx.doi.org.ezproxy.neu.edu/10.1377/hlthaff.2015.0755
Sommers, B. D., Gordon,
S., Somers, S., Ingram, C., & Epstein, A. M. (2014). Medicaid on the eve of
expansion: A survey of state medicaid officials on the affordable care act. American Journal of Law
& Medicine, 40(2-3),
253-279. doi:10.1177/009885881404000206
Thomson, K. (2012).
State-run insurance exchanges in federal healthcare reform: A case study in
dysfunctional federalism. American
Journal of Law & Medicine, 38(2-3),
548-569. doi:10.1177/009885881203800212
Vistnes, J. P., &
Cohen, J. W. (2016). Gaining coverage in 2014: New estimates of marketplace and
medicaid transitions. Health
Affairs; Chevy Chase, 35(10),
1825-1829. doi://dx.doi.org.ezproxy.neu.edu/10.1377/hlthaff.2016.0500