Responding to Response to Intervention

(Editorial Note — See also follow-up post here)

Like many well-meaning instructional reforms, Response to Intervention (RTI) is likely to fail if it is not coupled with other reforms that address the perverse incentives blocking its proper implementation.

The idea behind RTI is that we could avoid placing many students in special education if only we provided them with well-designed instructional approaches in the early grades.  The huge increase in special education enrollments consists almost entirely of growth in Specific Learning Disability (SLD), which is an ambiguous category that is difficult for practitioners to diagnose properly.  Almost any student with a normal range IQ but sub-par achievement could be labeled as SLD.  But of course, students may lag in their achievement because they have been poorly taught, not because they have a problem processing information, as is characteristic of a true SLD.  Schools have a variety of incentives to discount the former explanation and instead push students into special ed.

RTI is a federally-backed program that attempts to address this problem by allowing schools to divert 15% of their special education money into well-designed instructional programs for the early grades.  If students are taught well, they won’t be lagging academically and so will not end up being identified as disabled.

This all sounds great, but it is almost certainly doomed to failure if we do not also address why schools were not previously providing well-designed instruction in early grades or why they are so motivated to identify students as disabled.  Essentially, RTI frees-up money to get schools to do what they presumably should have been doing already — providing well-designed instruction in the early grades.  Unless we think that the main impediment to well-designed instruction was that schools lacked the funding to do it, diverting 15% of special education money to early-grade instruction will not get them to do anything significantly different from what they were already doing.  Even if we thought that the problem was that schools were unaware of the effective approaches that RTI offers, we have no reason to believe that schools will truly adopt or effectively implement those strategies. 

It is a a seductive but entirely mistaken reform approach to believe that schools are eagerly awaiting to be told by the federal government or philanthropists how to teach effectively but are just lacking the critical resources and knowledge to do it.  Schools already hire certified professionals who have been exposed to countless hours of pre-service and in-service training.  Why would we think that the only reason that they are failing to employ an effective technique is because they are unaware of it?  And with school budgets increasing every year, why would we think that the next bit of money is the one that they finally need to pursue effective strategies?

Instead, we have to recognize that educators have reasons for doing what they are doing.  They generally believe that the techniques they’ve adopted are effective, even if they aren’t.  Getting them to switch to something else takes more than just offering it to them.  This is especially the case when they’ve seen untold failed instructional fads come their way.  They’ve learned to tuck their heads down and do what they think works based on their own limited experience and inertia. 

RTI does nothing to address these barriers to instructional reform.  In addition, it does nothing to address the incentives that schools have to place students in special education.  In most states schools receive additional funding when a student is identified as disabled.  If a student is lagging academically and the school would have to devote some resources to helping that student catch-up, the school could either choose to say “my bad” and pay for those extra resources out of their existing budget, or they could say that the student is disabled and get additional money to help that student catch-up.  Of course, they have strong financial incentives to choose the latter explanation.  Research that I’ve done with Greg Forster and that Julie Cullen at UC San Diego has done, confirms that these positive financial incentives play a large role in the growth of special education.  That is, special education is growing, in large part, because we reward schools financially for increasing their special ed enrollment.

I know that many people claim that special education is a horrible financial burden on schools because it costs far more than the subsidies they receive.  But people who say this are either simply advocating for more subsidies or don’t properly understand what a “cost” is.  A cost is an expenditure that one would not otherwise make.  Simply showing that more is spent on special education students than subsidies received does not prove that the subsidy is less than the cost of identifying a student as disabled.  More is spent on students lagging academically whether they are identified as disabled or not. 

The positive financial incentive for identifying students as disabled exists when the subsidy is greater than the expenditure required by the special ed label beyond what would have been spent on that student anyway.  Because proper accounting is almost entirely absent in education, it is difficult to measure these additional costs directly.  But from the research showing the response to financial incentives, we know that there is often a financial reward for putting students in special education.

I don’t mean to suggest that educators are cynically gaming the school finance system or are even aware of its details.  My point is that the systems that school districts have adopted for the evaluation and identification of disabilities are shaped by these financial incentives so that even well-meaning practitioners will tend to over-identify disabilities when there are financial rewards for doing so.

Of course, RTI does nothing to address these financial incentives for increasing special ed enrollments.  In fact, it may contribute to those perverse incentives because schools are rewarded even more by placing more students in special education because they now get to divert 15% of that money for general education, which is essentially fungible.  And to make matters worse, diverting 15% of special education money away from disabled students may short-change truly disabled students who need those resources.

I’m sure that the people backing RTI are completely sincere in their confidence that we could prevent disabilities (and save money) if only we had proper instruction.  But wishing does not make that happen.  Reformers need to stay focused on combining promising instructional reforms with fixing the perverse incentive systems that undermine those instructional approaches. 

The incentive reforms should include changing the process by which we provide financial subsidies so that there are not strong rewards for over-identification of disabilities.  One way to do that is to provide vouchers for students with disabilities equal to the full value of what is spent on them in public schools.  That way schools would have to think twice before identifying a student as disabled.  Sure, they’ll get extra resources if they put a kid in special ed, but they also risk having that student walk out the door with all of his or her resources.  It places a check on perverse financial incentives. 

RTI with special ed vouchers could be a winning combination.  RTI by itself is just increasing federal subsidies for the status quo.

28 Responses to Responding to Response to Intervention

  1. matthewladner says:

    I think your point about incentives is broadly valid. There is however a school district here in Arizona (Vail) which put in universal screening and have had very good success with it. Their special education rates overall have dropped by 30% and their minority special education rates have dropped by 50%.

    I believe one of the students of one of Dr. Lyon’s colleagues helped put in the system.

    Incentives always lurk in the background, however. The Arizona special ed funding formula provides a very, very small amount for SLD kids.

  2. Greg Forster says:

    Here is my response to Matt’s response to Jay’s “Responding to Response to Intervention”:

    I agree. 🙂

  3. AwayWeGo! says:

    I’m glad you posted this. You raise a lot of good points, especially regarding students who are not succeeding because teachers are not examining and altering their own unsuccessful instructional methodologies. Perhaps the way we assess and support our teachers needs to change.

    Also, when you discuss providing vouchers for special education students my big concern as a parent of two special needs children is that with vouchers schools might not be so willing to work with parents who are trying to secure appropriate services for their children. It might be easier for them to just let the parent walk away with the voucher and send the child somewhere else, especially if the parent or child was known as difficult. But a shift in programs could be detrimental to the child if neither program adequately meets their needs and as a result they are constantly jumping out of the pot and into the fire. Besides what would the district really care if they lost some of the money from if a difficult special needs child left they could always make some of the lost money back by classifying a few extra students eligible for special education who might otherwise have had an unfunded 504 Plan.

  4. Greg Forster says:

    Your concern is one that a lot of people share, but the evidence indicates that private schools very much want to serve disabled students. For example, Florida’s voucher program for disabled students has grown to become one of the two largest voucher programs in the nation. Jay and I did a study on it a while back that found the participants were getting much better services than they had received in their previous public schools – and the voucher population roughly mirrored the statewide disabled population on a variety of indicators, meaning there was no sign of discrimination based on race, income, disability type, or disability severity.

    For a full treatment of this question, see the chapter on it in the Education Myths book.

  5. I don’t disagree with Matt (or Greg) that schools wishing to control specail education enrollments and make appropriate diagnoses would do well to consider RTI. My point is that offering RTI does nothing to get schools to take these steps if they were not already inclined.

  6. I should also mention that Amber Winkler at Flypaper had an interesting post on special ed last week ( )

    But as I emailed to her: “I think the views of beltway ed reformers on special ed are distorted by the exceptional circumstances in DC and among elite beltway parents. [Your] post last week is a clear example of this as are Andy Rotherham’s writings on the topic.

    It’s true that in DC there are some very sophisticated and aggressive parents and a very weak school district, making special ed a vehicle for savvy folks to extract desired services and benefits from public resources. But the evidence clearly shows DC as an exception to the national pattern.

    In the vast majority of cases special ed placement is driven by schools, not parents. And over-identification of disabilities is being caused almost entirely by schools, not parents. Schools do this, in part, because they are financially rewarded for placing more students in special ed. Schools get other benefits, like shifting blame to a disability from what could be educational malpractice, reducing everyone’s expectations so that lower outcomes are demanded of schools, and getting ammunition to lobby for additional funds.

    Over-identification is a school phenomenon, not a parent phenomenon. Parents are mostly compliant and deferential to school authorities. Even if parents wanted to “force” an unwilling school to identify a disability, they have to pursue a legal process in which they are highly disadvantaged and very likely to lose (according to Perry Zirkel at Temple University). And if parents drove over-identification, we wouldn’t see the relationship between financial incentives and special ed enrollments…”

  7. Greg Forster says:

    Just to be clear, I didn’t take Matt’s post to be talking about RTI. He mentions “screening” kids with diagnoses, which I took to mean “screening them to see whether they’re really disabled.” In other words, set up a special-ed auditing system.

    For the record, I agree with all of Jay’s analysis here, as anyone who’s read Education Myths knows.

  8. Amber has a new post on special ed over at Flypaper ( ). I disagree that medical or environmental factors have significantly increased special ed enrollments and made the case in a comment on her post.

  9. I’m not sure whether what Vail did is significantly different from RTI or not. My understanding is that they tested all kids at in the early grades and provided remedial help for those who were behind in their basic skills:

    This is very DOABLE, and if it is true that there are massive amounts of funds being shifted from general ed into special ed, there is an incentive for an enlightened administration to pursue it. I mean, for crying out loud, if we can’t at least agree to get good diagnostic practices in place….

    Having said all that, I still agree with Jay’s overall point. Kids win when you align the incentives of the adults to match those of the students. Inertia is huge in the system, and the sloppy and unscientific, and error prone methods and perverse incentives behind the identification of kids into special education today are just short of criminal.

  10. Greg Forster says:

    Matt, if there’s an incentive to do it, why isn’t it being done now? The incentive is NOT to do it, because it reduces special ed funding. You’re right that “enlightened” administrations have every reason to do it, but that’s not the same thing as talking about whether schools have an “incentive” to do it.

  11. AwayWeGo! says:

    Greg thank you for letting me know where I can get additional information. I’ll have to check that out.

    Also, from my experience as the team leader of a Child Study Team, I agree with Jay that over-identification is a school phenomenon, not a parent phenomenon. Based on my experiences over-identification is in part a direct result of schools not appropriately utilizing a full bank of available resources and strategies that would have prevented some children from requiring special education services in the first place. But there is also a lack of parental involvement and cultural factors that causes students to look like they have a learning disability when in fact they are really just environmentally deprived. They still get classified anyway because schools don’t know how else to help them or feel that it is the only way to cover themselves under NCLB.

  12. Greg-

    The words “if” and “enlightened” were chosen carefully. 🙂

  13. Greg Forster says:

    I see – you’re saying that if it were true that special ed was a burden on schools, they would be testing kids early to prevent them from being mislabeled. I misunderstood. But the trouble with THAT thesis is that if special ed were a burden on schools, there would be no need to test because the kids we’re talking about wouldn’t be diagnosed in the first place. 🙂

  14. This is really murky territory. My gut instinct is that schools actually do shift money from general ed into special ed for certain types of disabilities, but it probably depends on the level of effort that the school or district makes and the type of disability. I think public schools are revenue rather than “profit” driven, so if they get a good hunk of money for an SLD label but don’t do much more in the way of service, you are in perverse incentive territory.

    In Arizona, you get a very small funding increase for each SLD kid, making it profoundly rational to stop misidentifying children into SLD. After all, you get a very small amount of money, and the IEP process is a headache, and you will have to shift money out of general ed for most of all services provided.

    Notice also though that Arizona still has a fair share of SLD labels, despite the funding formula. Much of this must be just inertia and ignorance.

    At the opposite end, where schools get a fair amount of money, one has to ask: what is the incentive to do universal screening, assuming that schools are very concerned with maximizing revenue?

  15. Greg Forster says:

    You’re definitely right that schools are affected by “total revenue” financial incentives rather than “profit” financial incentives, as their behavior shows in many ways. In most states, the situation you’re describing as “perverse incentive territory” is what in fact prevails on the ground.

    You point out that Arizona has a lot of SLD kids. But it’s the growth rate, not the number, that we’re looking at. Jay and I found that in the dozen or so states that have reformed their funding systems, special ed is growing a lot more slowly.

    Interesting note: in our study, we listed Arizona as an unreformed “bounty” state, but you describe it as a reformed “lump-sum” state. Do you happen to know if Arizona has reformed its special ed funding system in the past few years?

  16. I think it is still a bounty state, it’s just that the amount of money given for SLD is very, very low.

  17. […] Greene is dubious about Response To Intervention — trying to educate children well so they’re not […]

  18. Kevin says:

    Montana is implementing RTI pilot projects. No one is diverting 15% of their funds. Montana’s school funding formula also has no reward for having additional kids in special education. Sped funding is based on the statewide average percentage of kids in special education. The amount every district gets for special education is for X% of their student population being special education students. The X% is the same for everyone, whethere their actual special education population is X+7% or X-7%.

  19. Greg Forster says:

    Right, that’s what Jay and I called “lump-sum” funding in our study. We were examining the difference in the growth of special education enrollments in states with lump-sum funding versus states with the traditional funding system that financially rewards schools for additional diagnoses (referred to by several of the state special education people we spoke to as “the bounty system”). We controlled for the presence of accountability testing in the state, since many people attribute the growth of special education enrollments to schools trying to avoid accountability testing. We found no relationship between the presence of accountability testing and the growth of special education enrollment, but we did find that almost two thirds of that growth was associated with the presence of bounty funding.

    See here:

  20. I don’t have access to the NBER working papers, but the abstract on the Cullen paper says “My central estimates imply that fiscal incentives can explain over 35 percent of the recent growth in student disability rates in Texas,” and “The magnitude of the institutional response varies by district size and enrollment concentration, student race/ethnicity composition, and the level of fiscal constraint.”

    I think that your initial characterization — “suggests that schools have financial incentives to increase their special ed rolls” — is fair, but that “special ed is growing, in large part, because we reward schools financially for identifying more students as disabled” may be an overstatement of Cullen’s findings.

    You also wrote that you did not “mean to suggest that educators…are cynically gaming the school finance system or are even aware of its details” and that “the systems that school districts have adopted for the evaluation and identification of disabilities are shaped by these financial incentives so that even well-meaning practitioners will tend to over-identify disabilities when there are financial rewards for doing so.” If I read this correctly you are saying that the system is shaped by an awareness of the incentives that “educators” don’t know exist (the same educators who often believe that special education creates a net district level financial loss). It doesn’t make sense to attribute actions to incentives – real or not – that are outside of or contrary to the beliefs of the actors. In order for the supposed benefits to become an incentive, they have to be recognized.

    Incentives come in many forms. I think it is unlikely that teachers and others initiating and handling special education referrals (including parents) are motivated by any possible financial benefits to the district as a whole, if only because, correctly or incorrectly, most believe that the state money accompanying a special education designation does not cover the costs associated with that classification. I do however think that many are motivated to initiate referrals because they see students struggling and know that one way to get services to a child is via special education. Special education also enjoys legal protections that make it relatively immune to budget cuts. Other interventions don’t have this protected status. I propose a different read on the institutional/systematic dimensions of classification and even over classification, one not based directly on funding but on a desire to get some children extra services via whatever means are available. Interventions outside of special education are often seen, correctly and incorrectly, as less available, consistent or thorough. If these perceptions are true, then there do need to be systematic and institutional changes to correct this. RTI is probably not sufficient, but it may help.

    This brings up another problem with the logic of your case. As I understand it the assertion that special education supplemental funding is a net gain for school districts is based on the assumption that “The positive financial incentive for identifying students as disabled exists when the subsidy is greater than the expenditure required by the special ed label beyond what would have been spent on that student anyway.” You note that we don’t have the accounting to prove or disprove this. You also write “If a student is lagging academically and the school would have to devote some resources to helping that student catch-up, the school could either choose to say “my bad” and pay for those extra resources out of their existing budget.” However, there is an implication throughout that districts are not providing these extra services outside of special education, “out of their existing budgets.” I would guess that the availability and costs of non-special education interventions vary widely, but we don’t know. The “what would have been spent on that students anyway” is an open question. As long as this is open, the question of the net financial benefits or harm of special education classifications is also open.

    Greg Forster’s comment about “total revenue” incentives as opposed to “profit” incentives introduces a different thesis. That argument gives no weight to the money that would be spent on struggling students outside of special education. Districts do have incentives to increase total revenues, but they also have pressures that work against all spending increases, particularly those targeted to what many see as “other people’s children.” Superintendents from around Wisconsin who I have worked with on school finance reform have opposed revenue cap increases tied to high needs/high cost special education classifications because they know that many taxpayers in their districts resent the money already spent on these students. In light of their rejection of this total revenue enhancement, I’d like to see Forster’s thesis more developed with evidence.

    I do know that the Madison, WI district where I live — based on a belief that special education classifications have a net financial disincentive for the district and red flags raised by disproportionate minority referrals and classifications — has worked to reduce referrals and classifications (see here:

  21. Greg Forster says:

    I’ll let Jay respond to your other comments on his original post, but I want to interject something where you say this:

    …”special ed is growing, in large part, because we reward schools financially for identifying more students as disabled” may be an overstatement of Cullen’s findings…

    But Jay was referencing not only Cullen’s study, but also the study he did with me. Our study found that the portion of special ed growth attributable to financial incentives was not 35 percent, but 62 percent. And given that Cullen’s study is a localized in one school district, whereas ours is a nationwide study, I think it’s fair to use our finding rather than Cullen’s to estimate the size of the effect.

    As for your response to my comment, you’re not really disagreeing with me that schools have incentives to maximize revenue, you’re pointing out that schools can also have other incentives that counteract those. And verily they can! But the question then becomes which of these incentives is usually stronger. The statistical evidence pretty consistently finds that, on average, schools are more likely to maximize revenue than not. Obviously that doesn’t mean there are no particular times and places where the opposite happens; it only means that on the whole, across large numbers of schools over long periods of time, the tendency is to maximize revenue.

    If you want to see this “more developed with evidence,” you can start with the Education Myths book; chapters 2 and 15 are relevant to this question.

  22. Thanks for responding.

    I should have been clearer on the Cullen, my point was that your study and findings and Cullen’s are very different and that lumping them together obscures these differences. “Large” is subjective. Most people would agree 62% is large; “over 35%,” is borderline large to many. It also seems that Cullen looked at other explanatory variables and from what I can tell from the link your study only looked at testing and funding structures as possibilities (please correct me if I am wrong about this). This isn’t a criticism of the analysis (you are free to include or not include whatever you want), just a desire to raise an awareness of what is and is not behind the word “confirm.” Policy discussions – especially on blogs — often suffer from inexact characterizations of research findings. We are all guilty of this. It can be awkward to include all the caveats and explanations instead of just linking, but sometimes it is worth it.

    No, I wasn’t disagreeing about the reality of total revenue incentives and I hope it didn’t read that way, just noting that my experiences suggest that special education may be a one of those places where the opposite happens. It may not. That’s why I’d like to see the total revenue thesis as an alternative to the profit thesis more developed with evidence in relation to special education funding. I don’t see that in the chapters you mention.

  23. Thanks to Thomas J. Mertz for his thoughtful comments. Let me respond to a few of his points that Greg did not address above. First, incentives do not have to be consciously recognized to have effects. People are often unaware or not entirely honest with themselves or others about why they do what they do. For example, tax deductions for charitable giving clearly increase total philanthropy, but very few people would say that they gave because there was a tax break.

    Second, you provide excellent examples of how districts can establish procedures and policies that respond to the financial incentives that individual educators do not recognize as flowing from those incentives. For example, a district might arrange it so that educators can only get money to help struggling students by classifying them as disabled. The district may do that because they know they’ll get a subsidy. The fact that the district does not make funds available to schools otherwise is a perfect example of steering educators toward identifying disabilities.

    Third, the lack of reliable data on costs does not undermine the argument that financial incentives are encouraging disability classifications. We see evidence of the incentive operating in the people’s behavior. That is, we look at what they do, not what they say their motives are. And we can look at the result without knowing every step along the way. Greg and I compared special ed growth in states with different funding systems. In states that rewarded additional diagnoses with additional funding, special ed enrollments grew much faster than in states that did not. And in the few states that switched systems to cease rewarding identification, enrollment rates slowed or reversed. So, we can see evidence of the incentive system by comparing outcomes under different funding arrangements. Julie Cullen also found that a change in funding incentives produced significant changes in special education enrollments, but you are right that she did not examine the issue nationally.

    Fourth, you (and I thought it was Matt) may be right that schools act as revenue maximizers rather than profit maximizers. If that’s true, then my argument is even stronger, because any subsidy will encourage over-identification of disabilities even if the school loses money.

  24. Greg Forster says:

    Picking up the question of whether schools are revenue maximizers: I think I wasn’t getting your point earlier. I thought you were questioning whether schools were “maximizers” at all, and the evidence I pointed to establishes that they’re maximizing something financial. But if you’re asking for evidence on whether they’re “profit” maximizers or “revenue” maximizers – understanding that here we are using both these terms in a loose metaphorical sense – that’s a different story.

    Looking back at the comment where I initially said that they’re revenue maximizers, I think I may have misunderstood what was implied by the idea of schools as “profit” maximizers. After all, the public school system doesn’t really have anything that’s unambiguously analogous to profit. All the money that comes in gets spent on something; there are no shareholders to pay dividends to.

    In special education, schools seek to maximize special ed enrollments, but it seems to be fairly broadly acknowledged that in IEP meetings they seek to minimize their commitment to deliver services. I can see how that could be interpreted as “profit” maximizing.

  25. […] response to the special education discussions (here and here) on Jay Greene’s blog, reading researcher Reid Lyon expands the discussion. Many […]

  26. Lynn Romanek says:

    I have a problem believing that schools deliberately set ut to “get more money” from the government and that is what propels labelling students as needing “special education”. Perhaps I am naive, but “those extra resources” as you call them aren’t going to benefit the general student population, nor the administration, they are going to be used for the special education student. At the same time I was intrigued by your conclusion that special education is growing because we reward schools financially for increasing their special education enrollment. I would argue that perhaps special education is growing because this generation of students have more challenges than earlier generations with respect to attention span and obtaining and processing information. I wonder if the advent of computers, computer games and internet access has not diminished the attention span of this generation of students and whether the trend will continue into successive generations.

  27. Susan Goding says:

    My problem with rti is that it is focused on reading and spelling, with basically complete disregard for content. There is no talk of intervening if a student struggles with science, for instance. The popularity of rti seems to me is that it reinforces the teacher centered model and so is a program that can be bally-hooed while making sure the structure of twentieth century education is unchallenged.

    This intervention will not increase differentiation, will not increase higher order thinking, and will do nothing to increase the intellectual capacity of students. Most of the curricula for these interventions are based on nonsense sentences, ‘the mouse moved the red house’ or ‘apples acting acrobatically’. Rti looks more like schools duplicating their failures, using the same failed interventions for rti as they have used for special ed students.

    If we want to bring these students along intellectually while they learn reading and spelling we should give students access to Universal Design for Learning and multimedia assessments and learning aids and differentiated instruction.

    Unfortunately, this would challenge the teacher led and the Carnegie unit models.
    This is not about the money; it is about the structure of education. Teacher centered interventions or student centered, differentiated, content driven interventions. The old way or something new.

  28. Naomi Boyle says:

    Has anybody heard of schools using federal IDEA/RTI funds for the purpose of “documenting” the need for a large percentage of the student body to be placed on 504 plans in locations where the state (not the federal government) reimburses school districts for 504 plan accomodations and modifications?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: