People Who Don’t Want to Collect Education Data Should Explain Why It’s a Bad Idea

Earlier in the week Reuters published a story about a $100 million education database being built by inBloom, a non-profit funded by the Gates Foundation and the Carnegie Corporation. The database will house data on children across the country, and it has the potential to help educators, researchers, policy makers, and yes, for-profit businesses. For the most part, the story is couched in concerns over the risks associated with privacy issues or the misuse of data. Yet despite the alarmist tone, there are no specifics that go beyond boilerplate quotations from fearful parents. Alexander Russo asks for more:

Are there problems with state databases being being hacked and releasing sensitive student data? Tell us about them. How do these issues compare to data security problems in general? How many states include Social Security data in their student records, and how does this compare with other public agency databases, which have their own Social Security problems? Once again, some context and comparative data would be more helpful than isolated data points suggestively linked together to convey fear.

The Hechinger Report also published a short piece on the database. Class Size Matters head honcho Leonie Haimson was tasked with making the case against the database, and her logic borders on incoherent:

“There are no limitations on the time-frame, or the kind of data. There’s no provision for parental consent or opt-out. The point is to give our kids’ data away for free, and share it as widely as possible with for-profit ventures to help them market and develop their learning products,” she says.

You may be asking yourself, “Is her argument really that we should stop the construction of data systems that could help millions of kids because we’re afraid it might make it easier for companies to create better educational products?” To answer your question, yes, it is. And it’s nothing more than mood affiliation. For-profit companies are bad, and therefore anything with the potential to help them must also be bad. In reality, there’s no reason this should be the case. Over the last few years I’ve requested and received “confidential data” from a number of schools and departments of education. It has always been for the purpose of doing research for a non-profit organization, but it’s conceivable that a for-profit company could use the same data to improve a product, and in that case it’s likely their use of the data would have a much bigger positive impact on American education than my use of it. So as long as the data is protected, why should I have access to it but not them? Perhaps Haimson is worried about these terrible things they might do (via the Reuters article):

CompassLearning will join two dozen technology companies at this week’s SXSWedu conference in demonstrating how they might mine the database to create custom products – educational games for students, lesson plans for teachers, progress reports for principals.

Progress Reports? The horror! The of end Haimson’s remarks is particularly telling.

”For-profit vendors are slavering right now at the prospect of being able to get their hands on this info. and market billions of dollars of worth of so-called solutions to our schools.”

What happens when “so-called solutions” are marketed to schools? Somebody at the school makes a decision about whether to buy them. There’s not some magic machine companies have where they input student data and a mind-control wave forces people to start buying their products. What happened to letting principals decide what’s best for their schools?

None of this is to say that privacy isn’t a legitimate concern, but there’s nothing new or exceptional here. At this moment thousands of school and government employees have access to confidential student data, and that data is protected by the same types of laws that will protect the data in the inBloom database.

Last night on Twitter Josh Barro said something insightful about Rand Paul’s fillibuster, but I think it applies here too:

I spend a lot of time with large sets of education data (both public and confidential), and so I’m admittedly biased toward believing the slope from “useful data sharing” to “gross violations of privacy” is not that slippery. Nevertheless, I think it’s important to investigate privacy issues and the weaknesses in data protection systems. That’s what keeps the slope unslippery. The problem with the arguments in the articles above is that they mention broad, hypothetical, worst-case scenarios instead of identifying specific problems. You can’t just yell “the slope is slippery,” you need to explain where and why the slope is slippery. Until critics come up with better reasons than “It makes me nervous” or “It might help a business be successful,” there’s no reason to believe the benefits of inBloom’s project won’t outweigh the costs.


Update: 3/9

In the comments Adam rightly takes me to task for glossing over parent concerns about consent, but there’s a reason I didn’t spend a lot of time on them. The debate over the database essentially boils down to the moral ambiguity of benevolent paternalism. My view is that the new database creates no additional risk — the same data is being entrusted to the same people who have the same intentions as always — and therefore collecting the data will help kids. That’s my moral reason for supporting this “regulation” on data collection.

Parents also object on moral grounds. They hate the database because something they didn’t explicitly agree to is happening and they feel violated. And that’s a perfectly legitimate reason for not wanting data to be collected. When it comes to the database there’s not one side that has the moral high ground. It’s like Bloomberg outlawing 20 oz drinks. You can’t say that either side was objectively “right.” One side fought for public health and the other for personal freedom.

The inBloom database has the same moral ambiguity. One side wants to improve schools and the other side doesn’t want data about their family taken from them. My complaint is that even if you have the moral defense of “it’s wrong,” you should still have a practical defense, particularly if the other side can also claim morality is on their side. For example, even if you think you think the Constitution gives you the right to buy a gigantic soda, you should still try to explain the specific downsides of a 16 oz limit. Hating the database because it’s unjust is perfectly logical, but the argument against the database would be stronger if there were specifics about new dangers.


Update #2: 3/9

There’s been some good discussion about this, and I want to reiterate that this is a complex issue that touches on real conflicts involving political philosophy, personal freedom, and public choice theory. That ambiguity regarding what’s “correct” is one reason why I’ve been somewhat vigilant in asking what additional practical reasons there are for opposing the database.

It’s also worth mentioning that the situation presents a potential collective action problem. For example, I think a fairly standard order of preferred outcomes might be as follows:

1. The database exists without my kid’s data.
2. The database exists with everybody’s data.
3. The database doesn’t exist.

The problem is that if everybody chooses option #1, you end up with option #3 and forgo the benefits of options #2. That doesn’t mean choosing option #1 is wrong, and every parent won’t have this set of preferences, but it’s something else to consider.


Update #3: 3/13

I’ll try this one more time because this post seems to be making its way around and a lot of people are having trouble engaging with the argument I’m making.

I am not objecting to the view that the database is terrible because it’s an obscene violation of privacy and a parent’s right to protect their child. If that’s your stance, there’s nothing more you need to say to make a serious and coherent argument. But given that there are people not convinced by that argument, I’m asking if there’s anything else. For example, regardless of whether or not Citizen’s United is constitutional, there’s a convincing case that the court’s decision has had a range of specific negative consequences. What I’m asking is, regardless of what you think about the legitimacy of inBloom’s right to have you kid’s data, what’s the case that their possession of the data will have specific negative consequences? That’s what Alexander Russo was getting at when he says “Are there problems with state databases being hacked and releasing sensitive student data? Tell us about them.” In the comments Molly says my invocation of “benevolent paternalism” should make the poor uneasy. And I bet when she wrote that she had in mind historical instances in which government overreach has hurt low-income families is specific ways. My question is what are the new specific ways the inBloom database will harm families?

Thus far, I’ve heard two responses. One ties complaints about broader trends in education reform to inBloom, but because they’re not specific to inBloom they don’t really address my question. The second response is based on the abstract notion that for-profits have difference incentives, are not well-intentioned,  and that this will harm schools. Is there evidence for this? Not amongst the many people who passed up much more lucrative jobs to work in education. In fact, the best evidence we do have, which is admittedly weak, is that over the last decade schools have made slow but steady NAEP progress, and this occurred during a time when the increasing presence of for-profits was supposedly destroying public education (not the mention the fact that economic growth has also been much slower in the last decade.)

So when I dismiss Leonie’s remarks and call them incoherent I’m not objecting to the idea that parents ought to be outraged by inBloom. I’m objecting to the fact that her remarks are in the context of explaining why it’s a problem that for-profits might get their hands on your kid’s data, but her argument essentially amounts to, “OMG for-profits might have your kids data!!”


17 Responses to People Who Don’t Want to Collect Education Data Should Explain Why It’s a Bad Idea

  1. Pingback: Remainders: In defense of Big (corporate) Data (about students) | GothamSchools

  2. the grey area says:

    One thing you may want to consider is that the companies you refer to spend incredible amounts of money on sales people and marketing. Once the starts to happen it makes it very challenging to figure out if tools are being purchased because they are useful, or for other reasons.

  3. adam says:

    Wow, you really not only misread parents’ concern and frustration over the sharing of their children’s data here, you also seem to be willfully ignoring it. The data was not collected with the intent to be shared for market interests, nor were parents who provide that information consulted when the decision was made to take such a significant leap in the use of such data.

    You need to look around a bit more, and actually consider others’ perspectives before poor strawman posts like this…

  4. Peg Tyre says:

    Great post. Data is gonna be great for schools. I mean, christ — teachers work by intuition most of the time. I agree, we can do better.

    But here’s what I worry about: I worry that private companies will get access to the inner workings of how children think and and how much they know. For example, your big multi-faceted corporation is collecting data on my low income cousin learn math. Turns out, at his under-resourced public school, the kids don’t learn math very well. In fact, my cousin doesn’t really solidly understand percentages and compound interest. In fact, he fails those question on the math subjects exams he must take by law to leave his terrible public high school — all duly entered into the data base. Oh — and by the way– that behavioral data collection part of his public school notes that he slow to develop impulse control.
    Now I ask you– HOW VALUABLE WILL THAT INFORMATION BE TO CREDIT CARD COMPANIES, SUB-PRIME MORTGAGE COMPANIES AND OTHER PREDATORY LENDERS? To be able to target people who don’t understand money — well, right now they do it by zip code and credit history. But big data collected at school will make that so much more efficient. And how long before they have it? And when they have it — what recourse do parents have? Who do they sue? I mean, the mom sent her child to school — as is required by law. And if schools are going to act en loco parentis AND collect and sell this data, I think we have to start talking about how to protect vulnerable populations of kids– poor kids, special ed kids, english language learners — from what are the natural and rapacious habits of corporation — who must function to serve their shareholders and bottom line.
    You see- we have two sets of interest here –and two big ways to use this data– ones that serve corporate interests and ones that serve civic interests. If you think they are the same, you do not know history.

    • Eric Horowitz says:


      Thanks for specifically bringing up a new danger that could arise from the inBloom project. That’s exactly the kind of thing I was asking for.

      My response is that in my mind this isn’t all that new. At any point in the last few years Capital One could request student level data from a school or department of education. Even if they came up with a better reason than “do better predatory lending,” I think it’s fair to say that their request would be denied. Now it’s true that the scope of who controls or has access has marginally expanded, but I don’t think there’s been a significant change — in a sense we’re still living by the rule of “trust well-intentioned people to protect kids’ data.” So my view is that even though on the surface there has been a change, it’s somewhat inconsistent to have been fine with the status quo last year, but now be afraid that new people who are not well-intentioned are going to have access to data on your kids.

      • Phillip Smith says:

        It is a significant change, if you factor in the ease of access. An example would be if I sent a letter to the wrong address. One person could possibly get whatever information I had in it. Now say I hit reply all, or send an e-mail to my entire address book by accident. Now hundreds may of received information that I meant to send to one. Now you say that you have gotten information from schools, but this would be on a whole different level. Instead of contacting the state, or district or school, you would be logging into a database built for ease of access and use. What might of taken longer, or actual contact now can be completed in a few clicks.

        “Are there problems with state databases being hacked and releasing sensitive student data? Tell us about them.”

        This would no longer be a state database, this would be a private database the states, schools, businesses, and individuals would send and receive information from. This sounds exactly like the way a credit bureau works. We also know that these are hacked or misappropriated all of the time. In fact, the first lady just had her information leaked the other day.

        I just feel that until some type of identity protection is instituted, the future potential for abuse is staggering.

  5. Tom Hoffman says:

    The question isn’t “should we collect educational data.” It is “should we create a single, privately-held database, backed by the full compulsory power of the state, to hold essentially all available educational data for all public school students?”

    If we do that, we’re creating a brand new, incredibly powerful institution with little discussion, oversight or accountability. Nobody knows where this goes next. The credit bureaus are the closest parallels I can think of (with fewer privacy protections than ed data gets), and they aren’t exactly shining beacons of awesomeness.

  6. Sorry you found my arguments “incoherent”; I find yours willfully blind. There are thousands of parents out there who find it outrageous that their children’s most confidential and sensitive data — including disciplinary & arrest records — is being shared with a corporation, whose sole purpose is to disclose it with for-profit vendors, and is doing all it can to encourage as many of them as possible to dream up products that would utilize this data, all in the name of “improving instruction”. All this without parental notification or consent. Perhaps the burden of proof should be up to those who want to utilize children’s data to explain why they are doing this and whose benefit it will serve? Only they haven’t told parents a thing. Perhaps the author should take a look at some of the comments on the Jeffco petition page here: or the Hechinger article to see how other parents feel. oh yes, “One side wants to improve schools and the other side doesn’t want data about their family taken from them. ” Sure, parents don’t care about improving schools, because after all, it is only their kids who attend them.

    • Eric Horowitz says:


      You’ve missed my point. I fully accept that there are parents out there who think this is outrageous. That’s a valid viewpoint and a perfectly good reason to oppose the inBloom database on its own. Again, I am not saying those parents are wrong or not understanding something. But this is clearly an issue where different people find different sides convincing. So given that split, it would behoove inBloom critics to give specific evidence in support of the claim that this new database will harm their children. Your quotes amount to an evidence-free claim that there is an army of evil for-profit companies who are going to harm your children by….well…I’m not exactly sure because that’s not explained beyond the vague notion of “marketing.”

      Of course there are always risks of security breaches. But once again, the arguments against the database aren’t presenting evidence that it involves a significant risk increase beyond the present risks we face from storing student data in a range of state, federal, and district databases. There are merely ad hominem appeals to the fact that “for-profits” are now involved, and they ignore the fact that the power to share data will remain in the hands of district officials, just as it always has.

      And I agree, it would be helpful if districts, researchers, and for-profits were more open about the ways they plan to use student data. One problem is that sometimes it’s hard to know before you see the data, but I do think it would be good to establish norms around making specific uses of the data more well-known.

      I also want to point out a few factual errors in your comment. In the 4th line you write “being shared with a corporation,” but according to the Reuters article inBloom is a “nonprofit.” You also wrote the “sole purpose” of inBloom is to share data with for-profit vendors when it is one of many purposes.

  7. Sorry, inBloom Inc. is a corporation, as I wrote (don’t you see that Inc. there?) There are for-profit Corporations and lots of non-profit corporations too. That it’s a non-profit doesn’t mean its goal, mission or activities are any less dangerous. And it’s primary purpose, as anyone would know who happened to read the inBloom website and that of its predecessor, the Shared Learning Collaborative, is to entice software developers into creating products that would access and take advantage of all this bounteous private data. That is why Sharren Bates etc. spends much of her time time traveling around the country, doing “learning camps” asking software developers for ideas and offering them prizes for the best ones. I have no doubt there are other goals that Gates has in mind with the data, including enabling states to share value-added info about their teachers, and making it difficult for teachers to get jobs if they move across state lines. None of this data, by the way , whether its your child special education or health conditions could be shared with vendors, without parental notification or consent, if either HIPAA or FTC rules applied. But if they happen to be in your child’s educational records? Or yes, “benevolent paternalism” applies. That is, parents are too poor or stupid to understand that this is in the best interests of their child, so it up to the Gates Foundation and/or their state ed dept. to decide for them.

  8. Shino says:

    Mr. Horowitz is obviously not a parent of a public school student in NYC. We have NO FAITH in the NYC Dept of Education, NYS Education Dept or for-profit corporations (or many non-profits for that matter) to do the right thing by our children. This debate has to be put in the context of what has been happening in NYC and elsewhere in the name of education reform. It’s called corporatization and privatization of our public education system.

    For-profit corporations are motivated by their bottom line: $. And, it seems our Education Depts are ever more willing to assist in that endeavor (so that top level people [e.g., Joel Klein] can get a cushy private sector job after they leave public service). After seeing the publishing companies flock to the Common Core State Standards, which I consider harmful to our children, so that they can capitalize on this new federal-state initiative, it is clear what these companies are after. Let me tell you they have no interest in improving our children’s education.

    So, in a different world in a different era we may not react the way we do. But given the current climate in public education, we have every reason to fear the worst.

  9. Lee says:

    Eric do you have children? Only a total idiot would ask this question. Lets publish all reporters personal data including address, SS#, income, email, DOB, test results in school, all your job history and income, your police record, your misprints over the years(including this one).
    Parent’s have the most to invested in their children’s lives. NCLB failed, it was full of data. RTTT will fail. Data doesn’t teach children but it does make the Edu vendors rich while stealing money from the schools and students. No amount of data will teach a child, only small classes and wrap around services.
    WAKE UP, you should be ashamed of printing this!

  10. Demetri says:

    I have two questions regarding all of this:

    1) what exact information is InBloom receiving from the participating states?

    2) InBloom CEO Ian Streichenberger is quoted as saying, “there is a concensus that personalized learning is the way to go” in education, but how does aggregating massive amounts of data about kids from all walks of life and synthesizing it into some lesson books etc actually help a teacher in a specific classroom inspire and teach a specific child?

    Perhaps i am too simple to understand all of this high tech data stuff, but teaching and learning has always been a simple organic process. It is a two way street. The student has to want to learn, and the teacher has to understand the individual student to be able to guide and draw that student into new knowledge and self awareness. Its a simple process, but it takes lots of time, effort and personalized dedication on the part of both the teacher and student.

    For all of inbloom’s good or bad intentions, i am not in a position to judge, i fail to understand when there is a “concensus” that “personalized” care is the key (and we know plenty of parents sending their kids to private schools who would agree that size does matter) why are we not spending more effort on making smaller class sizes a reality?

  11. Molly Sackler Wulkowicz says:

    When someone invokes “benevolent paternalism” as a positive thing, it should make people skeptical and deeply uneasy – especially women, people of colour , and the poor. Added to that, your dismissal of rightfully outraged parents and activists like Leonie Haimson as “alarmist” and “incoherent” makes it hard to believe a science writer in the 21st century would fall into such 19th/early 20th century postures of elitism and sexism. Just who are you writing to and for?

    As the parent of a public school child, I’m not fearful I’m furious. You want parents like me to articulate specific concerns? Well, we need specific information – or rather ANY information – if we are to make detailed arguments about why this is a criminal violation of our children’s rights.

    And if this stealthy, unauthorized stealing of our children’s private information is a good thing, why are only public school children so blessed? Shouldn’t children in private schools also have their confidential information taken, stored, and shared with all those benevolent corporate entities so eager to help children?

  12. mark phelan says:

    $100,000,000 boondoggle to take money that could be spent on something.. basketballs, teachers, pencils… and give it away to crooks in the hope it will open the door to other crooks to take more… what could go wrong?

  13. Jeff Nichols says:

    There is a larger context that concerns many, but far from all, parents, which is the general trend towards privatizing public education. For me the issue is the way that trend is manifesting itself in our society: through the interventions of politicians who accept campaign donations from the very corporations that profit from this privatization — politicians, moreover, who almost uniformly send their own kids to private schools. Process matters. This is the problem with supposedly benign paternalism: sooner or later it always becomes malign. Democracy is always better in the long run, and our schools are no longer governed democratically — too many decisions are made over the heads of parents, educators and democratically elected local school boards, by high-level politicians working with fabulously wealthy individuals who have unlimited resources for political and commercial advertising. So between the self-removal of the majority of the nation’s economic elite from the public school system, and the susceptibility of millions of the nation’s poor, who are dependent on public schools, to the blandishments of wealthy corporate and foundation heads, the level of public awareness of and resistance to the take-over of what in my view is a public good — the public schools — by private, for-profit entities is insufficient. Whether Bill Gates and Michael Bloomberg mean well is beside the point. And so is the issue of profit-taking in education. I have no problem with companies producing educational products for profit — as long as I am not forced against my will to purchase them. This goes for InBloom as well. If local school boards, with the full involvement of parents and educators, freely choose to participate in a form of data sharing that is making a private corporation money, fine by me. But it is NOT fine for that to take place because politicians whose own children go to private schools and who accept campaign donations from that same corporation make the decision over the heads of parents and teachers. There is simply no way I will consent to anyone having access to my children’s educational data until I as a parent, voter and taxpayer again have the right to participate in the decision making process — which in new York City, right now, I do not.

  14. Businesses have an obligation to their shareholders to maximize profit. Their attitude towards their customers is “caveat emptor.” Changing schools from a public service to a money-making opportunity, will benefit the investors, not the children. Moss’s recent book about the food industry is a good example of how private businesses can profit by disregarding the welfare of customers.
    So far the business approaches to public education have not benefited children. We should not be giving businesses access to any information about children. A friend of mine described her work in advertising as “identifying the soft underbelly of the consumer. This data base has been explicitly established as a business venture. We should not expose our children to these vultures.