participating in aco efforts in both public and private sectors. and now we're going to hear from a couple of the experts on looking at the bigger picture across all of the aco experience in the u.s. so far. next up for that i'd like to turn to michael mcwilliams. michael? >> so i'll speak more from a research and policy perspective as mark said. and i'll comment on three things. two challenges, and one as far as results go. and i think one of the key challenges ahead from a policy perspective is getting the benchmarking methodology right, the spending targets for acos. under the current model, the incentives for acos to generate savings are quite weak. particularly because of the rebasing that's implied by the current rules. so if an aco achieves spending on one contract period, carrying the current rules forward that would mean that their spending target or benchmark for the following contract period would be lower by that amount.

so to give you a sense of why that really diminishes the incentives, if you think about an aco increasing spending during a contract period, medicare shared savings program, aco for example that isn't facing any downside risk, they are penalized for doing that. and it increases their benchmark for the following contract period and then they can receive a shared savings bonus for doing nothing. under the pioneer program, they are penalized for doing more, but again there's that offsetting effect by the benchmark going up and the subsequent contract period, so for pioneer acos the incentives are more akin to fee for service with a lag. then, thinking about the fixed costs of investing in systems to actually control spending, the rebasing is that much more of a problem because it may be hard for acos to recoup their investments of investing in the right systems. so that is one challenge. and i know sean and his team is working hard on a revamped proposed rule for the shared savings program. another challenge that's related is unless acos have more control over where patients get care in the aco programs that diminishes

the incentives further. to give you a sense of that we analyze outpatient care patterns at baseline among acos and found, for example, that only 80% of patients that were assigned to one aco in one year were assigned to the same aco in the next. that's just over a two-year period. among the high-cost patients that number was even lower, about 75%. so that instability, that churn in the assigned population, suggests diminished returns on investing, in specific patients. a lot of specialty care leakage, so even among the most specialty oriented acos we found that over half the specialty office visits were occurring outside of the acos. and then something we termed contract penetration, the proportion of outpatient care revenue that's devoted to the patients under the risk contract, that was only 40% in our study. so that suggests very weak

incentives to roll out or implement changes that might spill over onto other medicare patients. and we think that those spillovers are probably likely, for example we found spillovers in massachusetts from the alternative quality contract which is a commercial aco contract sponsored by blue cross blue shield onto the medicare population. then lastly -- so those are two challenges that require some rule changes if not some developments in the market, like medigap select plans oriented around acos to help sharpen the incentives. and then thirdly, despite those weak incentives and the constraints that acos face the results have been quite positive. i think there have been some fairly convincing reports of early savings. we have some results coming out soon, demonstrating a positive impact of the medicare aco programs on patient experiences, including improved or more

timely access to care, better overall care ratings among the complex patients that are more likely to be targeted by the care management programs. as well as patients perceiving their care to be more -- better coordinated. so from my perspective i think overall it seems like a good start. but the need for some changes, new developments, to make the program a true win-win both for medicare providers, as well as patients. >> just one before going on to david, just one quick follow-up, is that stated as a resource hypothesis i think you're basically saying that if the financial -- the new financial incentives created by an aco program are pretty weak because of things like turn or shared savings, and no downside risk and the potential for losing out on that contract. and the subsequent year, you're not going to see very big effects. you have done some work, as you mentioned with the aqc, the blue cross alternative quality contract, which is, i would

think, bigger, in terms of those kinds of financial incentive measures, there have been bigger effects there, are you willing to say anything about whether doing more in terms of these kinds of strengths of incentives will lead to bigger effects, is there evidence for that? >> so the aqc doesn't have that rebasing involved. so if an aco is looking over a longer time horizon the incentives are certainly stronger. i think that's true among most of the commercial aco contracting. so there's a negotiated budget, and it will stay there, or there's at least no rule saying it will get ratcheted down every time savings are achieved. the aqc is rather interesting in that it was implemented in a country broad hmo network. so there are no restrictions, or financial incentives for patients in the aqc to go to the participating aqc provider groups. there are, however, mechanisms, because it is an hmo plan for providers to deny care outside,

because they do need -- they need to have primary care doctors, and the pcps can approve or deny referrals. anecdotally it sounds like the aqc groups, the provider groups have successfully without those financial incentives been able to corral care in and contain it within their provider groups. many of those groups are quite large relative -- >> and more effect on savings as well as quality? >> right. and so the other from the aqc is real savings effects and real improvements in quality. so whether that's related to the stronger incentives or not i think just remains to be seen. it's massachusetts. so, that's also a potential difference. >> still need more research. and i know david you've been doing a lot of work around the country on this. really appreciate your perspectives on what the research has shown to date. >> so i'm going to focus on three different areas. the first is the variability among acos. so when we talk about accountable care it's often referred as somewhat of a

homogenous group of providers, but, in fact there's a lot of variability among them. there are acos that have 30 physicians and there are acos that have 30 hospitals. as you can imagine they have very different needs, concerns and approaches to managing a population. and so what we're starting to see is that when you talk about accountable care it needs to be subdivided into the different aspects of the providers that are participating. and the reason for that is because they really have a different glide path. they have a different opportunity to progress toward achieving the same common objective. so if we want to provide better care, we want better patient satisfaction, and lower costs, it can be achieved, but with different provider types they're going to focus on different things. and they really should prioritize different things from day one. so that's one of the challenges, is that many of the acos are focusing on topic "a," when really given their individual structure they should be focusing on topic "b." the second area talks with the real challenge of becoming an aco and managing populations. there are two kind of core broad

groups of acos. some of those have been in effect acos for a long time. they've been managing populations, they have relatively integrated services. they have somewhat developed h.i.t., and then there are those organizations that have been strictly fee for service shops in the past and they're trying to make that transition. what we found is that it's a long and hard transition to become a population focused provider group. so it's not something that we expect organizations to make that transition over the course of the first year. it's not even something that's necessarily going to happen within three years. it's going to be an ongoing challenge. some of the big challenges that they mention are obviously the h.i.t., what they should invest in, and when. but probably the biggest challenge is just giving the provider buy-in. so you can imagine if you have a group of physicians that have been working on a fee for service basis for the past 20 years and you say now we need to focus on a population, getting them to change their practice patterns, getting them to work more as a team, getting them simply to change the referral practices, is a challenge and it

takes a lot of time and effort. and so while we'd like to see results in the first year and have a good indication of whether acos are going to work or not, one year really isn't enough time before trying to evaluate the organizations that are trying to make that transition. the third thing to talk about is the strengths of the different types of accountable care programs. so medicare, medicaid, and the commercial. medicare what i really view as an enabling program. it has a relatively low threshold so providers are able to go in and start to focus on population health in the short-term. also, if you only have -- if there's no downside risk, there's not a huge financial barrier or disincentive to enroll. and so while there are start-up costs, it's really a way to enable a lot of providers to start to bear risk and experiment with this. on the commercial side, we think of this as a program that allows additional resources to the acos. a lot of the large commercial

payers out there have a series of different contracts that they work with providers on. so initially they're not in a full risk bearing contract. they're not a full-blown aco. but they'll do pay for performance bonuses but it's a stepwise progression. where over a period of years they help train the providers in the skill set necessary to manage a population. and so it's really a longer-term track as opposed to just jumping in and being an aco right away. on the medicaid side, this is really where i see there's a lot of opportunity for states to push the accountable care movement. and the reason for that is that they have a disproportionate ability to force providers into bearing risk. and so where we've seen these states that have a strong focus on accountable care, we're seeing a lot of activity both within the medicaid space, and also outside of the medicaid space. as providers are being forced to consider becoming an aco for the medicaid program, they'll want to look for other opportunities to experiment, and so those where they're afraid that the medicaid program will move that way, the providers, for example, are more willing to just enter

into the medicare program. the last thing that i would make is that aco growth within a market has not happened by itself. there's very much a lot of strong market dynamics at play. and you don't see an isolated aco. when there's one aco all of the competitors in the market either form an aco right away, or create an aco plan over the next few years. and so, within individual markets, there's a lot of overlap in terms of the different providers that are participating, and it really is happening at a market level. not just at the organizational level. >> david, thanks for summarizing a lot of evidence and experience quickly. you know, i just follow up on the point about states and medicaid program. you mentioned a lot of activity, including some fairly significant reforms, away from fee for service payment. anything you can say about actual results from some of those programs so far? they are pretty early. >> sure. the one good example is oregon. they've, what we would consider

to be most aggressive tactic in moving their entire medicaid population to ccos, or care coordination organizations. and they have seen generally positive results. they are mixed. but they are somewhat positive. there are a number of other states that are just starting this year to move their population or next year even or the next few years, and so it really preliminary results even earlier than the shared savings program. >> i'd like to thank both of you for those research presentations and maybe go back to marcus, and our other michael, to get a follow-up sense from you about whether that overall summary of a national experience with acos today fits with your own. you all are obviously not at that early basic stage of just starting to get provider buy-in, and making some of the initial investments. you are both organizations that have been at this kind of approach to patient focused care for awhile. but you're different kinds of systems. advocate has a hospital base involved. and marcus, you highlighted that

brown and toland is a physician group. any comments on what you heard? or points that you'd like to especially emphasize or clarify? >> yeah. i think particularly that point about the fee for service transition, and it being a painful one, spending a lot of time in california, i think sometimes, you get a little bit isolated, we're surrounded by groups that have been in capitation, managed care, heavily so for a long time. and then when i have the opportunity to come out of the state, and meet folks who really have a managed risk, and you really begin to get a sense for what that really means. you don't have any of the infrastructure that any typical organization provider group that i would encounter in california, in other states where there's just fee for service, or heavily fee for service, they never had to adapt. they've never had to develop

these services. and so, beyond what happens at the provider mentality level, by the organization that's supporting them, there's a lot of work to be done, and that takes a lot of time, a lot of investment. but one point i do want to make, which i think is important, as a physician, one opportunity that i hope is not squandered in this, is that the fee for service churn is so disaligned for the patient, and the provider. and there really is an opportunity here where folks can get the right care with the right person at the right time. physicians, if they're willing to give up aly bit of their autonomy, and work in a team-based approach, will be able to spend more of their time working to the top of their license, and most importantly, spend a lot of time talking to the patients who really need them the most. i mean that's what really drives

me. what really motivates me to see this work through. i'm fortunate that we've got a great group of physicians in our provider group, and i think that more than anything has to do with our success. but that's what i really worked towards. >> just like to delve on that. david's comments on what does it take to change culture of an organization. so you've heard, you know, some examples shared in terms of physician culture. we, too, while we've had ten years of experience working with our physicians to improve quality to our clinical integration model, it's been a shift for our organization to now take a perspective outside of the four walls of our hospital, and to do that we created what we call an advocate care index, population based measures that all of our senior executives down to the director level, as well as in the physician side, create alignment for the organization to move towards improving overall quality and cost of care regardless of where that care is delivered.

and the physician office, you know, the points about patients supporting staff to practice at the top of their license, a cultural shift for our organization, you know, has been actually physically placing care managers that we hire into independent physician offices. so that's a big shift in dynamics, in that office space environment. >> and back to michael's comments about the shift in your financial support to make those kinds of business models sustainable. i expect there's some challenges around especially being a hospital-based system where a lot of the revenues have come from the kind of procedures that michael is highlighting, is maybe getting that revenue getting in the way of making some of these changes. how have you dealt with that? >> yeah, so, it's -- clearly the

model needs to move towards taking cost out of the system. and that is through, i think there was a comment earlier, about evidence-based practice, reducing unnecessary use, so that's a component of it. there's also a component of how do we achieve greater economies. and, you know, part of that scale in terms of an organization, but also looking at greater efficiencies in terms of systematic change, marketing services, human resources, those types of aspects, supply chain. so we're on a three-year journey to take out about 2% of operating costs each year to make sure that the model is sustainable. >> i'd like to open this up to discussion from the audience, as well. so we have a question in the back there. we'll bring the microphone over.

>> josh sideman, from the advelair health. for payment delivery and innovation. dr. mcwilliams it was really interesting the comments you were making about the beneficiaries experience of care. i'm curious if you are actually also the providers from the groups, have any insight into whether some of the incentives in the sometimes critiqued cms approach to beneficiary attribution and assignment are having an impact on how you're engaging beneficiaries. >> so it's hard to know from the empirical analyses, we've done what's specifically mediating the changes in patient experiences that we observed. as far as the mechanisms for patient engagement go, i doubt they're playing a big role. my understanding is that most medicare beneficiaries are unaware that they're even in an aco. they get this letter about data sharing that some are confused

by, and some throw away, and some opt out of. but the patient experiences that are in the quality part of the aco contract are very much more along the lines of caps measures. overall ratings of care, and physicians, and how access to care is. whether physicians are interpersonally connecting with them. whether their care is coordinated. those are things that i think the acos are probably impacting more through the systems that they're putting into place, whether it's new scheduling referral, access systems, care management programs that are really focusing on the needs of the complex patients. rather than anything going on between medicare and the patients. now, this does beg the question now how will patients view the medicare aco programs? i think certainly one concern is that if patients do not like the brand of care that acos are providing, then that would not bode well for the programs because they are voluntary and

that could lead to market share losses and maybe discourage groups from staying in the programs. >> michael, marcus, how much do your patients know that they're part of accountable care arrangements, or are they just experiencing care differently for other reasons? >> yeah, i think it definitely varies, and there's truth in that there are beneficiaries who probably don't realize that they're in the program. patient engagement is definitely a key element that needs to be executed on to be successful. and we also participate in some other cmmi innovation projects like iocp. so i've had a lot of discussions with other provider groups, and it's interesting. the medicare population is not one size fits all. that there is regionalism even down to the level of certain cities. so like for our group, reaching out, sort of cold call style to beneficiaries was not well

received. whereas in other communities, that was acceptable and they were able to get patient engagement. so we generally had to kind of circle back around to our providers and work through those offices to get to those patients, and i think to your comments, that as far as patient experience goes, i think that we have put a point of emphasis in the provider groups for patient satisfaction that get score cards every quarter. it's part of their bonus incentive program. so there has been some attention and some improvement there. but a large part is, as you were saying, you're creating a large organization behind the providers, and the patient experiences all of that. and that is generally good. so that's a good thing. that's coming out. even if it's not directly coming from medicare, or directly coming from the provider's office, per se, patient is still benefiting. >> and david, from your -- you highlighted the different types

of acos and how that might mean different steps that they should take to achieve some of the goals. presumably better patient experience is a key goal for all of them. any comments about differences that you're seeing, and how they're approaching the challenge of more effective patient -- patient engagement and experience? >> one observation is that i agree with michael that there's not really that focus on telling them that they're part of an aco, but it's really the benefit of the view by what are the providers serving. they say we're going to provide additional care coordination. we're going to provide enhanced access with primary care. but it's not saying you are part of an aco now. that's not the marketing strategy. across the groups it really does vary. and it also varies across organizations. some really don't put the -- excuse me, the patient engagement as a high priority. others are trying to build their entire strategy around that. so a lot of variability across organizations. >> thanks. next question, up here. and then in the back.

>> i'm dr. carolyn coughlin a primary care physician and an attorney, and this is actually a follow-up question. i'm perplexed by the fact that you, the acos, many of them are getting high satisfaction ratings, and you're also having trouble with churn. that people are drifting outside the aco. and usually sick patients who are happy with their care, what they're concerned about is being forced to change their provider, rather than -- they want to stick with providers who know them and that they know. so i would think that would reduce churn relative to a sort of a wide-open fee for service. and i'm curious about who is drifting away. and if you looked into why. >> i can address that. >> okay. >> so, what you're talking about is the attribution methodology,

and the joke i have is that attribution is the longest four-letter word in health care right now. it's an evolution in terms of how much data is being used, and into the methodology, and the logic. and i think that fundamentally is the issue at hand. and so, what we experience at pioneer over the first two years is about a 30% turnover rate. between people going out and people coming in. so, it's -- and i get, doctor, the question that you're asking. and it's just -- it's not as simple as that and it's not just about the relationship between the patient and the provider. it depends on, for instance, what doctors did you put into the attribution level. so for instance, we have 1700 physicians in our network. but only 300 of them are part of the pioneer aco. and that was a strategic tactical decision that we made. so that's one aspect of it. there were some rules in the

beginning about visits, how frequently they were occurring, and if they weren't occurring you would lose the attribution. so that plays a role, as well. so it's -- and actually we found that for folks who aren't technically attributed to us, they're still going to see our physicians. they're still connected in our network, and that's not going to be true for, say, orange county where there's tremendous amount of competition. more provider groups competing. the bay area, san francisco, is a little more concentrated. so there's a number of factors that go into it. and i wouldn't take what you're hearing as an indication of dissatisfaction from patients.

it's more about the attribution methodology. >> other comments on this? i know the attribution methodology and this issue of relationship to churn, and relationship with churn to beneficiary dissatisfaction is an important area. and all of you have thought about this. >> i might just add that if there is increased satisfaction, then it may very well decrease the churn. i don't think we know that. we don't have a good handle on that. but there may be a sizable effect from that. and the other -- i make one other point, which is that i think sometimes the knee-jerk response is to fix the attribution methodology. but that probably won't accomplish a whole lot without addressing the fact that beneficiaries assigned to an aco have unrestricted choice of providers without any incentives to direct them back to the aco providers and we just know there's tremendous care dispersion among beneficiaries in part for that reason. and that's sort of what the acos are up against. >> other comments on this?

okay, great. a question up here. in the front of the room. >> hi, morey menacker from hackensack. i heard sean discuss very, very briefly information, and his comment was that monthly they were supplying claims data to the acos. but we all know that the claims data from cms can be up to 12 months delayed. so even if you're getting it monthly, you can't accurately predict what's happening with your population and make changes. now, we've had some brief discussions about health information exchanges, and data sharing. but it seems clear to me that the successful organizations, be they acos or not, are able to share data, clinical data, and be able to make appropriate decisions based upon that

information. but yet, the government, with all its regulations has absolutely no regulatory oversight over electronic medical records and transfer of information. and has sort of left that to the private sector. the question is, is there a mechanism that we can utilize to improve data sharing to be able to really change our ability to quantify best practices and minimize wasteful spending. >> you think we're going to come back to policies, new policies around information technology and data sharing later. i would like to ask both michael and marcus emphasize the importance of information technology in everything that they're doing. but, what is the state of the research evidence on the extent to which better access to data,

whether it's through health information exchanges or other areas translates into better or greater success of the aco effort? or is that just another area where the research is limited? i'm asking david and michael to start. >> so, this is more observational data. the first is that, information technology is still very much a work in progress with most organizations. so, most of them have an emr that works. they're able to look at patient level data, being able to aggregate that to a population and track that over time is more challenging. there's a lot of investment right now going into predictive analytics and care management platforms. but whether or not they're choosing the right platform or doing it at the right time is still to be determined. so there's a lot of kind of push and pull among acos about how they're going about this. one other observation about information exchanges is that the primary focus of the organizations is sharing

information within the organization as opposed to sharing between organizations. the focus is trying to keep people in network and communicating between people that are quote/unquote on the same team. and so the focus right now is how do you make sure that your outpatient is talking to your in-patient. particularly when you might be on different emrs. we know of acos that have over 20 different emr platforms. they're just trying to integrate and speak amongst themselves, let alone trying to bring in all the other outside platforms. >> i would just add quickly it's a major research challenge. and i think that the data that david's group, for example, is collecting will be a major contribution to allow really rigorous analysis on that. >> and michael and marcus, any final thoughts on working effectively with cms data? >> yeah. we've spent a lot of time working with their actuaries trying to sort of solve the black box. i think one last closing comment i would want to make is, value-based purchasing is here. there was a lot of hand wringing

back with the election in the supreme court case, but by that point, particularly with deals like what devita did, it's very clear the marketplace has spoken. and folks that are the doing fee for service now have got to start changing their ways. because this is coming. it's in medicare fee for service. it's in commercial now, and so it is imperative for their survival really that they start making the changes we're talking about here today. >> any other final thoughts on the panel? >> it is with this past-moving area, and this fast-moving attention around value based payment it is a very challenging area for the research to keep up. and we very much appreciate the discussion that all of you had about the state of that research and its implications for further steps with acos. so i'd like to ask all of you to join me in thanking our panel for an excellent discussion. [ applause ] all right. we are now going to take a short break until about 10:45. we'll reconvene with our next panel on big issues for acos

going forward. as you can tell there are a number of them. thank you. we're going to move right into our next panel and while they are coming up to the stage, i'd like to talk about some of the big issues for acos, aco policy going forward. i would like to re-introduce the ton top ig of big issues for aco policy going forward. some of the main topics that we are going to discuss here build on the discussion that you have already heard. you have heard about the state of the evidence on accountable care and about some of the key issues and challenges facing accountable care organizations on the ground. now we are going to turn to the potential policy and regulatory reforms to help address the challenges. obviously some of the topics have come up today. but just to help get the discussion going i wanted to highlight a few. one of them is the issue of the

benchmark and payment systems. we heard about some challenges around how the benchmarks are calculated and whether the savings are retained over time or go into a new base. how attribution works. we have talked about transitioning to more person based payments, payments that are more at the person or capital tated level. that means more significant financial risk, but it's something that many acos seem ready to take on. we have talked extensively on the last panel about steps and policies that could support increased beneficiary engagement, issues related to performance measures have come up, both performance measures that are less burdensome to report, but also measures that could be more meaningful for capturing issues like underuse of care, and more meaningful patient reported outcomes and the like. we have heard about challenges

related to data availability since timely comprehensive data is really critical to many of the steps to improving care being targeted and having effective impact. we have talked about some of the other payment reforms going along with accountable care organization payments that can potentially reinforce their effects. we have talked about steps to overcome startup costs through things like bonus payments and other incentives. we have highlighted the challenges of implementing effective reforms in clinical care and how opportunities to share experiences and research on what really works to improve care and lower costs in particular kinds of health care settings and markets today can be very important. so plenty of opportunities for further discussion, debate and hopefully constructive next steps on the policies affecting

the aco programs in the country and particularly the medicare aco programs, very timely issue right now with the upcoming regulations and further payment and regulatory reforms that are coming. we have a great panel to discuss these issues. i would like to introduce them briefly now. starting on the far end is their president of heritage medical systems which is an affiliate of the heritage provider network that serves over a million patient members and integrated programs through medical groups and independent practice associations in california, new york and arizona. heritage provider network also manages one of the largest pioneer acos in the united states and one that seems to be doing quite well so far in the pioneer program. next i would like to introduce joe damore, vice president for

population health management at premiere. joe has extensive experience in leadership roles for successful hospitals and health systems including 19 years as a hospital and health systems ceo. he has successfully developed several integrated health systems and his expertise includes governance and leadership development, business planning, quality enhancement, health insurance plan and aco development and physician hospital integration. no shortage of skills needed for success in these kinds of health care reforms. next i would like to introduce my long time colleague mark miller, executive director of the medicare payment advisory committee. medicare payment advisory commission. medpac is a nonpartisan federal agency which advises the u.s. congress on medicare payment, quality and access issues.

mark came to medpac in 2002. has it been that long? where he previously was at the congressional budget office as assistant director of health and human resource division. before that mark miller was deputy director of health plans at the medicare program. last but not least i am pleased to introduce paul ginsburg. paul is chair of medicine and public policy he is university of southern california and a fellow here at brookings. from 1995 through the end of 2013 he was the president of the center for citing health system change and prior to finding health system change paul served as founding executive director of the physician payment review commission now part of the medicare payment advisory commission. as with our last panel we will start out with opening comments from our distinguished panelists and then turn to further discussion of some of the policy issues and potential solutions on the horizon. mark, let me start with you.

>> thank you. heritage is probably at the other end of the spectrum for many people in the audience in participating in the aco program. of our million patients 800,000 are fully delicated completely capitated. we pay claims and do care management and we have almost 3,000 physicians in medical groups that are employed. we probably have 30,000 contracted physicians primary care and specialty and independent practices that surround those. we try to focus on concentrating in markets and presenting a delivery system that can actually produce differentiated results. our success in the pioneer, i think, is primarily driven by the experience that was referenced earlier. dr. murken has been at this since the 70s and been through the ups and downs of different versions of managed care. we were able to extend in the pioneer aco program all of the

internal things that we do for those prepaid populations. so the independent practice that was a first timer to the aco gets the social worker who is going to work with the family, gets the community care worker, gets the meals on wheels. i was pleased to hear the conversation about consumers because we have about 30% churn. does that sound familiar out there? how do you financially survive that? what we are finding given we have had success and taken the education road with consumers and found they didn't know a thing about it, didn't understand they were in it, so we had group meetings not to market, they are already in the program, but we are a part of your physician office. if it is the patient or the family member who often accompanies them here is what we have added to your physician's office capability.

then when they get moved by whatever attribution mechanism they are calling back and saying why am i moved? i liked the social worker who was working with us and take care of mom or what have you. so that is a challenge, i think. the secret to our success like the hackensack system, we spend money that we make elsewhere to try to demonstrate that this program will work. it ultimately like all the organizations that are participating if we don't evolve the payment mechanism to something more population based it will run off the cliff. the big organizations like us or hospital systems, other big medical groups may have the staying power to stay with it a while longer. if you don't end up doing something other than like the

rural thing, effectively what was done when you give money in advance for providers who don't have it and then have to give it back out of your savings. that is reverse capitation. that is what we had in advance. we don't have argument about whether or not you need an air conditioner or whether or not you need a ramp. we don't have to look up the benefit design. the answer to access is could you come in this afternoon? if you can't come in and you don't have anybody to bring you we will go get you. if you can't come at all we will send somebody to see you. we think with all of those providers and hospital relationships we have we are great at taking care of people when they present and need care, but more importantly is can you find them before they present and they are sick enough or injured enough to fall in the door so maybe they don't need care or a different kind or not as intense? we think the program is a great

learning round. i think we have proven that with the independent practices. we want to see what happens next and be an advocate for the change that is coming. >> thanks. please go ahead, joe. >> i work with a team of people in population health management who work all over the country. we work in about 150 markets in the last 3 1/2 years helping organizations transform into population health on. on the medicare side we work with pioneers and issps. we work with commercial players to build new models. we work with medicaid programs putting in disparate programs across the country. so we work across the entire continuum. what i'm going to try to do is just summarize how our people feel about including maury's organization at hackensack. our organizations that we work with have done much better than

the national average in regard to hitting savings rates like hackensack. they have done a great job. we have a lot of others that have done well. i think almost 80% of the organizations we work with hit the minimum savings rate. in year one. not all of them got shared savings but a really high percent hit their minimum savings rate. we would like to think it is because the exhaust from the success we share with each other, we have a collaborative that allows us to share with organizations what is working and then the second thing we really try to do is stage what you build because what we find is many organizations try to do too many things too early. and what we have learned is stage them over time and there is a staging process that we think works. so our experience has been number one on the mssp and pioneer side, people need accurate timely data and we are

still not there yet. it is kind of like flying a 747 without an accurate instrument panel. if your attribution changes 30% in a quarter it is really difficult to get an accurate denominator on all of your quality metrics, your financial metrics. we have to fix that attribution issue in this next round. and then on the top line, if we don't have accurate claims and timely claims you really can't manage on a day in and day out basis. so we need to fix that. so data is really critical to be successful going forward. second, is there is a number of areas in the economics that need to change. for example, the minimum savings rate issue. as i said we have so many organizations that hit their minimum savings rate. what happens next year it is dialled back to zero again. and that doesn't make sense. so somehow that you should be able to take a credit for the success this year and carry it

over towards your target for next year. it seems terrible to say you did pretty good this year but you didn't hit the minimum savings rate but you are going to start all over again. that doesn't make sense. we think the change might give you more and more like that. the risk adjustment process is not really sound in our opinion. the benchmarking and the metrics need to be realistic in setting in their targets rather than saying you got to 100% of some of these metrics in order to get full shared savings. the third area would be where do you think we need to go. we think we need to allow organizations to continue the one sided risk model for several reasons. one is if they don't have totally accurate data i don't think it's appropriate to go to two-sided. that doesn't make sense. also there are organizations, and i think it was a great summary earlier about the types

of organizations in mssp. there is a lot of organizations that are in risk for the first time. i think it is too early for them to go to two-sided risk, is our experience. second, there are organizations that are ready to phase in to two-sided risk and they should be given the opportunity in that model to maybe go to track two and phase it in over time. the third is we do have a small number as mark pointed out earlier of organizations that want to go to full capitation or full global payment. they should be given that opportunity. the fourth we have a couple that would like to go into full capitation plus part d. we see those four options as a prudent way maybe to offer the current mssps the ability to stay involved, make decisions that fit their market and their organization. we are hoping that cms will come through with something like that. i hope that makes sense. >> joe, thanks for the comments. mark, i know medpac has extensively analyzed these

issues, and has a number of recommendations already, and i look forward to hearing from you. >> and i think there is some overlap with what you -- >> we have talked in the past. >> there is a couple of things. we spent the last year, year and a half, talking to acos, site visits, case studies, surveys, and lots of people who roll through the office to talk to us, and we put together a set of short-run comments which i will go through, and then we started talking about what the direction is for the future. there's two products, i have to mash it into five, seven minutes here, which i will do. but there's two products you should keep in mind. there's a june letter that we wrote to the administrator. obviously we're always writing to the hill staff. when we do this. that's up on our website. and then also our june 2014 report talks about some of the futures stuff. from the comment letter we were

trying to make comments based on things that we had heard from the field that we can also reconcile with the taxpayer and beneficiary interest on the medicare side. we made a series of recommendations for the next round of the acos. and on the issues of attribution and prospective bench mark -- or the bench mark which has already been mentioned, we made the argument that both of those need to be on very clear and consistent prospective basis. so whether it is set up to be retrospective, your attribution changes during the course of the year or the benchmark changes during the course of the year, or whether there are methodological issues that effectively change the benchmark as you're going forward, that shouldn't be the case. case. the benchmark and attribution should be on a prospective basis. there is a certain stability and just predictability that we think will allow the acos to manage better under those kinds

of circumstances. now, there is much more detail in how to do that in the letter but the basic issue is attribution and benchmark on a prospective basis. a legislative thought on the side is that for attribution purposes, advanced practice nurses, nurse practitioners p.a.'s, should be part of the attribution process. again, that requires a change in the law. and again i won't get into this because it's much more complicated than five minutes, but there is a specialty attribution process that we have talked about changing. we would say the aco should be able to designate certain specialists who are engaged in relatively primary care types of activities. you know like when you're going to your cardiologist because you have a heart condition and that type of thing. again, there is more detail but to blow by it in less than five minutes. the next thing is that i would say was a big deal in our comments is the quality indicators.

throughout this process we made the point that there were too many and that they weren't particularly the best ones, at least in our judgment, that should go. we were concerned about the administrative burden and actually just some of the burd and, actually, just some of the measurement issues. we're very clear early from the aco saying that this ended up being a lot more resource-intensive than they have an tis pated. and we have argued pretty consistently not just in acos but in ma and fee-for-service, but a smaller set of population-based measures. there's more detail and more complexity. again, you can look at the letter. we do want to encourage movement to two-sided risk. we know that there's some reluctance and concern about that, but we do want to move to two-sided risk. there's a couple of connected thoughts here. one is we heard a lot of concerns, and some of it was echoed in the previous section

and right here about the notion of being able to engage theorde leakage issue under control and issue the beneficiary in their care. if the aco is willing to accept two-sided risk, they should be able to forgive cost-sharing for the beneficiary. and this way, some of the way you explain the aco is rather than explain here's an aco, here's how it works, which lots of people, it's hard to understand, is to say, look, if you go to your primary care physician, there's no cost-sharing. you kind of get them in from that point and then people should know how to manage from that point. and then, the other thing we said is if you accept two-sided risk, there should be some sort of fee for relief. once the risk shifts to two-sided risk and some actor

says i'll 5:00 september the risk, then a lot of that should fall away. we gave examples we e where we, ourselves, should start focusing on regulatory relief to roll back. we heard lots of kmentss from the acos wanting to talk about enrollment model, a capitated model, which i think involves more discussion. and, also, on other ways to engage the beneficiary. and in wrapping up, the june 2014 report taulks about the future and how the aco sort of fits in a large picture. the point that we're kind of driving to is the notion of a single, unified benchmark that says for fee for service, aco and mma, the government may have a support rate. and then, within a given market, whichever model can dominate or be a strong player that model can come forward.

ma plans have a lot of overhead. they have enrollment, they process claims, they do that type of thing. and you may have a yutilization to finance that and still might be a market. in aco, you can capture one or two points and make that model relative to mma and fee for service. >> it's a lot to throw at you. >> mark, thanks for coming along and also giving people the references. paul? >> i'm going to avoid repeating what was said which i agree with, but let me point out that i see the con september as providers taking a moderate degree of performance risk. but could con septembers cannot overcome flawed details.

and the medicare aco models have been important catalysts through acos becoming a cig nef cant part of payments. and i think the short comings have been better understood over time and this is a time to fix as many of them as possible. my sentiment is that the acos have been tolerant as long as risk is one-sided. it's the kftsd in the model and whether they're actually willing to risk a loss in the model. so i think it's an affected aco model if it's most important there. which is medicare. so i won't repeat -- i'm in greemt about the comments about

patient attribution, beneficiary attribution that beneficiaries need to know, who they're responsible for and the ability to reach out and engage them. i had a couple of thoughts that mark started off about enrollment. i think in the long run, we need to develop an enrollment model aco. and, in fact, cms could offer temporary reduction in the part d premium for enrollment as an incentive to go there. an enrollment model would allow a network approach for models and higher cost form for others. it's a way of engaging more specialists in the aco, who may not be part of the risk sharing body, but, in a sense, could have a network relatirelationsh with the hmo. i think if you're going to go to

an enrollment process, you need to find one that's going to have a very low add min strative cost. as mark said, a lot lower medicare advantage. perhaps just involving beneficiaries with one from the aco that they are atriblted to of would you like to enroll? here are the benefits for you. it might be a way of getting into that model. i think i'd want to run the enrollment alongside the attribution model because of concern about the number of years it might take to get a critical mass of enrollees. and i wouldn't want to jeopardize the entire approach in that interim. i think many of the reforms that involve incentives are going to need medicare supplemental insurance. we know that medicare raises over on medicare spending and that overall reform in that has been overdue for decades.

but, basically, you cannot work with large proportions of beneficiaries having no point of service financial responsibleties. a group i work with, i have a bipartisan policy center and a 2013 report recommend that medigap be limited to providing catastrophic protection and needs to leaf half of the co-ininsurance and deductibles uncovered. this is particularly important to design any reform so that medigap does not cancel incentives for beneficiaries. i also have some comments on benchmarks. the benchmarks we have now are based ed provider-specific historical spending. i i this it's the right approach to get started with. it's not ideal, but we needed it -- when we have an

essentially voluntary program. otherwise, aco by aco against it. the uniformed national dollar amounts for updating are helpful in the worst short comings of this approach. we're coming up to a critical decision about benchmarks for second aco contracts. and i think rebasing to the aco's most recent experience would substantially undermine the business model for acos, which is not for trauma to start with. the long term beth to better benchmarks involves higher payments before -- oh, higher incentive to other providers to

work in acos. and the tricommittee built to fix sgr really takes this approach when it created atms or alternative payment mechanisms with strong alternatives. so if you have some broad incentives to go there, then you can start bringing in community or county-wide experience into your benchmark with acos. i think ultimately, we're just going to have to go in that direction. we should not write off the opportunity of legislative act for bigger changes. we certainly don't have the

greatest expectations addressing policy changes even when art san divides are not that pronounced. but we should take note of the far-reaching changes in medicare physician payment to fix the sgr that had unanimous support in all three committees of jurisdiction. i think a commitmented seen as incentives for alternative payment mechanisms could perhaps be brought over. obviously, given the sgr fix and you're waiving physicians the opportunity, you want to make sure those opportunities exist. it really means it should be natural for congress if it goes passed the lame duck in dealing with the sgr fix to start

thinking about critical changes to really assure that there will be real opportunities for physicians to engage in alternative payment mechanisms. thank you. >> paul, thank you very much. and i want to thank the whole panel. i went through a range of each individually challenging policy issue in introducing this panel. e i think you all hit on every single one of them and more. you know, we heard from the earlier panels about people who were engaged in doing aco work now about how in many ways, this is sort of like building the claim out while you're flying it. so there are lots of changes on the fly, during care delivery,

with existing policies in place. but, also, what came through is they're anticipating things changing. that there is because of market changes payment systems, employers, changes in states, things are going to be changing, similarly for medicare as you all focused on in many of your remarks. so it seems like on the policy side there's very much the same business. we're sort of building the plane out as we're flying it. hopefully, that refuelling will take place in the air when you need it. we could talk a lot more about the details, but i'd like to ask sort of the bigger picture than sort of where the planes are flying, too. mark, you touched on this -- sorry, too many marks on this panel, miller, touched on his

description of the long term vision that med-pack is led out and the bipartisan players had some similar kinds of long term reform recommendations, paul. but let me just ask you all, five, ten years from now, where would you like to see policies collectively get us in terms of support i supporting the next roupds of itch leaptation and the next roupds of care implementation. there are a lot of important policies to work out. >> i think population-based is the key, moving payment that direction again. we will have a bias towards the capitation side. just some folks are terribly

uncomfortable, moving away from fee-for-service, it screws it up. it's what we have to go hunting for. is this going to be paid for, whatever and we just tell our doctors ignore it. i think the enrollment idea is just perfect. these are old. if you would observe your business, you probably don't deserve the patient. and i think that's one of the things we want to create a set of incentives that makes sure that in every market possible, there are competitors. we don't want to be complacent as scientists or physicians or hopts or health plans or anything else. you want people vying for i can do that better for you. i would like you to trust me and

work with me as opposed to just when you're too sick and fall in the door. those types of things. >> we want progress. so every market has characteristics. but, in new york, in the new york area, when i went there in 2004, it was like rolling back the clock 20 years, whether it was benefit design or behavioral practices of providers. and the things that are routine are, like, you know, communistic takeover of parts of the greater new york markt. now, everybody is moving that direction because, as was pointed out earlier, the writings on the wall.

this movement is going to change this. based on having your benchmarks set based on your market other than national clinical standards, i think there's certain scientific things, the benchmarks ought to be market-based and you need to pay for progress. how did they roll up to a five versus seven? you can have segments of a plan, that contributes to those who operate at 87. around you're making progress on the whole book. let's focus on making progress and some of the graduate stuff that joe was talking about. but not too slow. >> great. i think you're spot on.

we call it the end game. and a lot of the organizations that we're in with would xiv e be to have a population base contract with medicare, medicaid, all the major commercials. so if i was the ceo of an integrated system, i might have 700,,00 people that i'm responsible for and i would problemly have a global target of expenditures for each group. and then it would be my bonus that might be based on quality, you know, patient satisfaction, engagement, health department status improvement. so that's kind of what we see as the end game. we are seeing a formation right now in a multi-number

populations. so they're coming together and forming a multiorganizational population. i hope that makes sense. >> i think for the commission, we've tried to be really clear. we're starting to frame up what the longer run looks like. again, these are comments that are subject to change. and, again, i think i probably garbled this, but you could view the world as this. you have fee-for-service and you're probably going to have a long time. by the way, there are parts of the country where it's a relatively efficient operation.

there's a relatively high set of overhead that goes with this. and then, on top of that, they get da profit margin. its add van fang is that it doesn't have overhead. they're not induring those costs. so, in a sense, maybe they perform well in markets where you don't have very high utilization, but you have some high utilization that an aco could extract a pointer to.

in some markets, those models will be mas, some it may be acos. there's a whole market there and for the moment, i'm going to lay that to the side. now, atribute in addition to having low overhead, has this element to it. in attribution, they can wonder around. so you have to keep the patient satisfied. that's an important element of what the contribution requires. you have an na program and one question is why would you

replicate that somewhere else. if you've benefitted, they're willing to actively make a decision. that might be a road that could be discussed. and the last thing i'll say, and he made this point, i think it's pretty important to keep that in mind and to make these critical, you know, make them workable models. if you had a strictly enrollment model, i'm not sure acos would be able to get enough model, dlooes currently. >> i've sensed that fee-for-service whether it's medicare advantage or acos or final payments, you know,

hopefully, will become more over time. i think the real challenge of getting there is both you have to put out a policy. and then i think the policy process has been appropriately attended to the experience of those engaged in the programs. i think probably the biggest challenge is the readiness of different organizations to go different steps of the way to wards population. the more just made about the multi-owner account organizations, that's a direction that i hope we see a lot more of. my concern is that so much in the way of scarce management resources, as well as money, is going into megamergers.

>> i'd like to open this up to questions from the audience. back in the back there? >> a couple of the panel iszs spoke about specialty positions. i was wondering if you could expand on that? >> some specialty physicians like cardiologists tend to be fairly involved in acos. but there are a lot of specialists whose care is not that intragal but, realitily,

the acos have no reason for that. i think the problem is the need for network acos to get the specialist more engaged. and i think some political needs to be involved in acos could happen with the fix that we've seen. >> i think one of the things that presents this evolution is big challenges for the organizations that have absorbed a lot of them. so if you're a big, multispecialty group or a hospital system, because of the way the system worked, your incentives were to collect these providers.

and make sure that they needed to do everything they needed to do. hospital systems in particular that are going to move in to accepting risk, all of the sudden, how long can they afford the big purchases that they make. they're very open about that. the model has changed. your estimates, maybe you sold bonds three years ago. your estimates of how many admissions, how many stents are you going to put in? it changes if you implement every best practice that we know of. we know that takes time, but it's a big, economic issue in terms of how do you change the system, take that cost out and make room for the very best providers to participate. >> in one large organization, they've decided to develop

models for three groups of physicians. one is the primary care and then the specialists, the sub specialists are really in-patient focused. and then there's the hybrid group in the middle who really do both. so they're looking at models to involve them more in managing their primary care loan. the second model that we've been working on is what i would call an episode of care model withms. they want to further engage specialists. so they've designed a a care program within an mms to look at setting a target and then sharing savings among the mmsp. so that's a model to further

engage specialists that maybe have not been as engaged in the past. so those are a couple of models that we're seeing evolve in markts throughout the country. >> for the second type, involving more discreet, specialized aspects of care, as opposed to cardiologists and end kronologists, heart failure management, diabetes management, things that are part and parcel to aco performance models. are there some other quality measures? >> i would imagine most people are flr with choosing wisely.

so we're seeing the adoption of choosing wisely for the appropriateness of specialty and subspecialty earnings in large, integrated -- clin integrated networks. i think appropriateness is coming into play in those subspecialty areas. >> one of the things that was -- that i was thinking about in particular, i've got to get this out before i forget, is as we experiment with respect to specialists, you may want to be careful how much we lock that in. if it ends up in law or a regulation that's so strongly defined you can be the get out of it, i mean, some of science is moving rapidly. so the same cardiologists are going to use an entirely different approach that doesn't

involve other kinds of treatments and it's going to eradicate. i'm sorry, your follow up question? >> there's a lot of cancers and rheumatoid arthritis and things that are not covered. >> this is really nerve-racking with all of these marks up here. okay. so a couple things that the commission has said and reacting to these comments. i think the commission's comments and is letter about the notion of letting aco designate certain specialists as attribution nodes and droektly consistent with what joe was say ing with the coordination of

care. these are very consistent with your first model. to your point, mark, on quality, i think there is real concern and this concerns not just the aco space but fee-for-service and elsewhere, that any time we bring a specialist in or bring in some other provider, we've got to bring in a specific specialist. i know a lot of people got it to say my thing.

a small set of quality measures that money can be awarded or not. and then the responsible organization can decide which message internally it wants to chase its particular model around as opposed to a single set of measures for all of that type of thing. very great concern on that part of the individual measures. we have something of a different view there. i almost thought you were making this point. i think you have to be very careful because some groups are

starting to wake up and say okay, you need to build an episode around my care. and in some ways, they see that as a way to preserve what they're getting currently. and i think we should all pay attention to that comment. i think those processes could be changing. >> i just want to say something about the quality for specialty care. i think what mark miller sketched out was probably fairly practical. but i don't want to give up on the fact that often specialty specific measures are much stronger than the general measure for a population. i was influenced by studying the joint replacement.

they seem, to me, to be much, much stronger than many of general measures for hospitals. so i think it's worthying about some very strong quality measures that may only apply to a very important frequent procedure and working into it at some point. >> time for maybe one more? >> you guys have all spoken about different changes to the aco program. i was wondering if you could speak a little bit about what extentd you think legislation is needed to make those changes?

number one, a long with a lot of people, i was going around talking to both sides of everything. it is right-minded and headed at the right provider to get more than others. so we need to get comfortedble with that. i believe it takes legislation to solve the payment changes that are going to go off of a cliff if we don't reset how things are paid. they won't have the legal authority from a regulatory standpoint to make a change. >> other thoughts on the ledge slative out look? >> well, i think a lot of the recommendations that we made in

the june letter can be achieved regulatorily. although there are things in the legislation that we tried to point out as we go. and then i tried to point out there probably is a legislative change that needs to go along with this. it's always clear in the congress says these are the changes i want. >> any final comments for the panel? thanks for covering short term. [ applause ] >> i'd like to thank all of you in participating for today's event. we'd like to work on these

issues in the future and obviously accountable care and the aco experience with some notable results on quality, some important successes on costs and medicare and medicaid and private and employer kinds of plans. obviously, a lot more work to do. we talked about the research on what can succeed in care transformation and public policies to support it. we've talked about some of the policy steps that can help people engage in care reform on the ground. both provide organizationsizations and patients and consumers do that more effectively. i just want to give a quick thanks which made much of this possible both in terms of the technical expertise, the contests around the country and people are actually engaged in aco activities.

for right now, i want to thank you all very much for attending and thank you for your concerns about improving the quality and addressing health departmentcare in the united states. have a good rest of the day. blapg blank

. >> we ark which youly had ebola right here in this country. and it's not a stereo. people are just reacting to a very serious disease where you have to have teams of doctors and nurses waiting on you constantly. but we had the isis fears, too. >> i think the people who are in the area of ebola should be deemed from entering our nation and it is up to our leaders to do that. >> i would like to see c-span do a question about is this ebola virus the proof that we need a

national, one-pair health departmentcare system. we just seen what happened in texas with this capitalistic health care system. that's to see what we get there out for remarks on that. that this is the proof that we need a national health care system. >> and continue to let us know what you think about the programs you're watching. >> you can send us a tweet at c-span hash tag comments. - >> with live coverage, here on c-span3, we comp limit that coverage by showing you the moegs relevant public affairs

events. >> a new report considers changes, the u.s. deportation system since 2003 and how those chaengs could affect future

policy. next, homeline security experts enforce this event hosted by the institute. this is an hour and a half. >> objection. we're going to get started with, of course, saying good morning and welcome to our program this morning. my name is doris misner. i'm a senior here at the policy institute where i direct our work on u.s. immigration policy programs and where we are very pleased this morning to be releasing a new report which we're calling deportation and discretion reviewing if record and options for change. now, this event this morning is in person for those in the room

and also live-streamed. so we welcome both sets of all yepss. the report that we're going to be talking about is coming at an important moment in that there is, again, the possibility of an executive action on immigration, sometime towards the end of this year. and that executive action may well take into acount some changes in the guidelines that present presently are being used. this work offers the most detailed look yet at the record of removals and deportations from the department of homeland security since its inception. so what we're looking at today is data from 2003 through f.y. 2013. the work builds on a large body of work that the migration

policy institute has developed over the years that have to do with various aspects of the enforcement system. this particular work today deepens a report that we released in the spring called the deportation dilemma. it digs more fully because we had more data into the information and into the findings. that we developed at that time:

who is the director of the homeland security and justice team at gao, the u.s. government accountability office. our second comment will come from an enforcement fellow at the immigration council and works extensively on the criminal justice system and immigration connections. so, with that, i'm going to turn the mic over to mark. thank you, and, again, welcome. >> thanks to you at mti to help pull the group together. i want to especially acknowledge my co-worker who was unable to be here today. so, this report analyzes every formal removal imp lemted by dhs

between 2003 and 2013. basically, it gives a profile of who's being deported from the u.s., how people are being removed and where they're being removed. we spent some time talking about how the actual enforcement practices line up with dhs's announced priorities in 2010 and 2011. and then we explore how potential changes could affect future removals. lets me just briefly say something about how bedid this. the main thing that we did is to analyze add min strative records from isa's data base, which is their main data base. the data was obtained by the new york times.

isa's data base includes a certain number of informal returns. our report and spent trags today focus on formal removals. that's basically cbp removing to mexico and canada. and then, throughout the report and my presentation, the data measure removal events, not unique individuals. so what that means is some people have reported more than once so it's a bias in these sort of repeat customers. all together, 3.7 million people have been removed. so that's the universe that we're looking at over that 11-year period.

just sort of a basic profile, 91% of removals have been men versus about half of all immigrants who are men. 91% have been to mexico and the three north earn triangle countries. voer sus about three quarters of unaut rised countries. so this system leans in those directions. this is lifetime, has anybody been convicted of a crime. most people, 59% of all removals since 2003 are of people who have never been convicted of a crime.

this was especially announced when removals of non-criminals grew faster than criminals. so the share peaked at 67 prnt 342008. so that drove the sheriff noncriminals down to about 50%. you can see where the lines converge there. and then, more recently, removals of noncriminals have, once again, grown faster. so in 2013, about 55% were noncriminals. the data and the report allows to get into much more detail about the types of crimes that people have been convicted of, people who have been removed have been convicted of.

the right side is just looking at zooming in on 2013. so, as i mentioned, 59% have never been convicted of a crime. and then looking at the specific types of crimes, about 13% of people removed have been convicted of a violent crime, and then the remainder are other violent crimes and drug crimes. to kusing in on 2013, the share of immigration crimes has gone up. the other wedges are all within

about 1% of their decade-long average chlts so one of the key findings of this report is looking at the aggregate data hides some real interesting images between people at the border and people apprehended within the united states. what you can see here are a majority of immigration-related crime. those two categories together, have consistently accounted for 82 and 90% of border removals throughout this whole period and about 80% in 2013. among people who have been

convicted, that represents about 4%. the picture is completely different for interior removals. for the entire period, about a third of everybody who's been reported within the united states were non-criminals. that number fell to 20%. so that means that three quarters have been convicted of a non-immigration crime.

so looking at the left side of the picture, you know, as i said, about three quarts of people removed from the border have nevada been convict of a crime. among interior removals, about a quarter have never been convicted of a crime. that's the first two edges. 28% have been convicted of violent crimes. 15% are in a category of general, nonviolent crimes, 18% are drug crimes. and 13% are traffic offenses. of the traffic offenses, about two-thirds of those are d.u.i. so, again, very different pictures at the border and in the interior.

one reason it's important to understand the difference is where enforcement is taking place. the first thing that you can see from this picture is at record levels during the obama administration. the lowest year -- there's some confusion in the popular discussion about removals. but there's no ambiguity that removals are at an all-time high. but looking a little deeper, we

see four distinct periods. so in the first couple yeefrs, removals were basically flat. and then what we see beginning in 2006 is a steep increase between 2006 and 2009 in the total number of removals. what happened then is with the failure of the comprehensive immigration reform in the senate in 2006, the bush administration is releasing a new surge. those numbers flattened out during the first years of the obama administration. we see the total number of

approvals was sort of flat and that there's no real change in the ratio of border to interior enforcement. most recently, we've scene removals that have started to increase again and then we see increased border enforcement. so very disht things going on. one other way removals differ is through the pipeline. the majority of removals are expedited removals. another 31 ppt are reinstate h statements of removals.

in the interior, we see over half of all removals are judicial removals. and it's -- we see that many are expedited removals, as we would expect. oat it's noteworthy that it's a nonform of judicial removal for certain criminal offenders. so one last sort of removal pattern set of informational i'll share. looking at the time, as we would expect, people are apprehended much more quickly at the border.

the pattern is is different for interior. it's a way of seeing that these populations look different from each other in terms of their fixativeness in the u.s. and it's also interesting to look at this reinstatement because this is a somewhat controversial procedure. if they get ream ri hended, that previous order gets reinstated without having an opportunity to appear before a judge and maybe argue that your circumstances have dhanged. what we see is 72 weeks. so the other key set of findings or another key set of findings from this report, and, you know, key sort of set of analysis is that we compare and the bottom

line finding is that most removals are consistent with dhs enforcement priorities. this falls outside of dhs's current priorities, the bottom stripe, the purple striet. and then people who have been con vekted of crimes. let me make three-points about the overall finds. 95% of all dhs removals fall within three or more of these categories. so 95% of everyone removed meets one or more of dhs's priorities. and the second point includes 93% of everyone removed by the bush administration. so the bush administration adhered to obama's enforcement priorities. so when you look at these data,

it's a little bit surprising that the moral ton memos should have been a major source of c controver controversy, both historically and in the data, that they formalized existed, you know, going back to the i.n.s. era. having said that, a third observation about this slide is that we do see some of the trends that i talked about before also translate in to adherence to these priorities. so, although the pry -- morton memos didn't break radical new ground they did sort of redirect enforcement and what you saw is that during the bush administration's enforcement surge between 2006 and 2008, and we lack at this in more detail in the report, the share of nonpriority cases increased, and actually 2008 was the year that nonpriority cases peaked at 10% of all removals, and that trend reversed as soon as obama came into office and the

nonpriorities fell to less than 1% in 2013. and you see that especially after the memos came out in 2010 and 2012011. so the morton memos have been effectively implemented but they've also been criticized by some advocates for american rights because they define dhs's priorities very broadly. so in particular, i mentioned the three categories, recent entrants. recent entrants are defined by dhs to mean anyone who is apprehended within three years of entering the united states as well as anyone apprehended by cvp regardless of the time line. immigration obstructionists are defined as anyone who's ever been ordered deported -- excuse me anyone who's ever been ordered deported as well as anyone who fails to appear at immigration hearing. and criminals are defined as anyone who's ever been convicted of any crime. so those are all sort of a broad

definition of those terms. so what this table does is two things. one thing, is in this first row what we look at is if dhs were to strictly adhere to enforcement priorities meaning anyone who fell outside of those would not be deported how would that change the numbers? this table only looks at the ievlt c.e. removal cases. we can't particular lee cvp cases here. what we can see here is removals would have fallen from 2.9 million to about 2.7 million. a 7% reduction. what the rest of the table does is ask the same question, but, if the priorities were defined somewhat more narrowly. so instead of all criminals, what if we only focused on people convicted of nonimmigration crimes. or what if we only focused on people who have been convicted of violent crimes. the way to read this table is to

basically look at these numbers. excluding immigration crimes from the priorities doesn't actually move the needle. i mean when you look at in detail it only changes the removal numbers by 7,000 over the entire eleven-year period and the -- i'll say something about the reason in a moment. so, we don't really see these numbers go up until we look at really significantly narrowing the criminal categories that we're going to focus on by not emphasizing level three i.c.e. offenders or violent criminals. all of these other numbers are actually smaller than we expected. so that was something that we were interested in and we looked into and one of the reasons for that, is that most people who fall into one of i.c.e.'s priority categories fall into more than one of the categories. so, the example i was just about to say something about the immigration crimes, most people convicted of an immigration crime are apprehended at the border and convicted of an immigration crime so they're not just seen as an immigration

crime priority they're also seen as a recent border crosser. most people reinstated are apprehended at the border and may also have been convicted of an immigration crime so a lot of people fall into all three priority categories or fall into two of the three priority categories. and the obama years, 58% of all removals fell into more than one pry orty category. so that's one reason why the numbers in that previous table were a little smaller than we expected. so in this table we looked at some more complicated scenarios where, well what if we change more than one of these priorities at the same time? what you can see is this does show that removals would fall more, you know, which is what we'd expect. and i'll just sort of talk you through a couple of cases, because it's a lot of numbers to throw at you. so, what this row is showing is that if dhs no longer focused on nonviolent criminals as a top enforcement priority, and they no longer focused on 10-year-old

yes row moveal orders, didn't consider people who had been removed ten years earlier we would project removals would have fallen by 17% over the previous eleven years. what this row is showing you is if dhs no longer emphasized non-dui traffic crimes, and defined recent entrants to mean be who came in in the last one year instead of three years removals would have fallen by 9% versus 7% or 8%. the most sort of far reaching set of changes that we model is to de-emphasize nonviolent criminals to or deprioritize to deprioritize people who entered more than a year ago, and people with 10-year-old or older removal orders and that would have reduced removals by about 19%. these numbers are bigger than the previous table. one of our takeaways is still that we were surprised that

playing with the enforcement priorities or doing these -- thinking about the enforcement priorities doesn't produce a larger reduction in numbers, and so one reason is because most removals -- a second reason is that most removals fall into core dhs interests. most people removed from the bored are apprehended at the border, within three days of entering, and those are people that dhs views as an enforcement priority. most people move from the interior have been convicted either of violent crimes or of drug crimes or dui, and those are people who are viewed as sort of core enforcement priorities. so for that reason, change in the enforcement priorities doesn't push the numbers down that much. this relates to a third point. which is that, if we think about exercising prosecutorial discretion during the enforcement process, as a form of executive action designed to reduce removal numbers, it's important

to recognize that discretion during the enforcement process has a different kind of impact on immigrant communities than an affirm to doca style program in which unauthorized immigrants would actively apply for relief. discretion during enforcement is less concrete and it's more difficult for immigrant communities to observe because most people who could potentially benefit from discretion during enforcement don't ever experience that benefit because they never enter the dhs enforcement system. so if you think about an unauthorized immigrant who lives in the united states and never commits a crime, they just don't interact with dhs for to the most part and are never going to see dhs's discretion in action. so, that's how this type of discretion should work. it means that i.c.e. is focusing on high priority cases instead of, you know, these people who would benefit from discretion. it also means that the psych local impact of discretion during enforcement and the politics of any changes

to enforcement priority are quite different than those for a an application based doca style program where people would affirmatively seek relief. so i'm going to pause there. i'll just remind you my three key takeaways, stark difference between border and interior enforcement across the board. the vast majority of current removals meet dhs's enforcement priorities and changes to the priorities would likely have a modest impact on the total removal numbers. >> okay. all right. okay. gives you a good summation of an enormous amount of numbers. more of which you will find in the report itself but they actually do really add up to these broad conclusions. so with that let me turn to you, rebecca. >> thank you. doris. and thank you, mark. and good morning to everyone. i just want to thank mpi for inviting me to participate in today's discussion. i really appreciate the opportunity to be able to discuss some of the gao's work

looking at some of the issues and topics related to some of the issues and topics that are raised in mpi's report which are being released today. before i turn to my remarks on some of gao's more specific reports i just want to give you a little bit more background on me and also who gao is in case some of you may not be as familiar about who we are as an agency. so as doris mentioned, i'm a director in gao's homeland and security justice team, and in that capacity i lead our work an border security and immigration issues. gao as an agency is located in the legislative branch. we are sometimes referred to as the investigative arm of congress or the congressional watchdog. and so in that capacity we examine how federal taxpayer dollars are spent and we examine how government programs and initiatives and efforts operate.

will of our work is done at the request of chairs or ranking members of committees of jurisdiction in congress, or our work can also be mandated by committee reports or public laws. our reports are available on our website which is gao.gov. and many of our reports contain recommendations to agencies that we are reviewing on changes or improvements that they can make to their programs. on the border security and immigration work in particular we have covered a range of topics. on the border security side we've looked at issues like goals and metrics for assessing border security efforts. with also done work looking at dhs' management and oversight of efforts to acquire and deploy surveillance technologies along the border. we've also done work looking at how to agencies th h