Recovery.gov Description FAIL

by

picture-6

As is now showing up all over the social web and news reports, the site Recovery.gov – which was set up to help the public hold government accountable for stimulus spending – lists a stimulus contract awarded to Clougherty Packing LLC for $1.19M for the listed service of “2 POUND FROZEN HAM SLICED.”

Read as stated, this clearly leads one to believe that the U.S. government spent nearly $1.2M for 2 pounds of ham …sliced.

Now, as has since been clarified via a press release from the U.S Department of Agriculture, “2 POUND” refers to the individual size of each of the 760,000 pounds of ham that were actually purchased – not just one very very expensive ham.

In putting out this press release, the USDA has completely missed the point.

The point of setting up and hosting a website such as Recovery.gov is to allow the public – and in particular, journalists – to look through it and hold the government accountable for how an historically-massive amount of money is being spent. When launched, one expects such a site will contain accurate information that is backed by the government – especially if the government is going to spend $18M dollars on building said website. Without accurate and understandable information, we as citizens and journalists simply can’t do our jobs, and the consequence of us not doing our jobs is potentially (even likely) irresponsible spending of $787 billion.

No laughing matter.

The stimulus is something that impacts each and every American’s life, and the fact that a website is set up to allow any ol’ citizen to see how their money is being spent is a laudable effort. At least it’s laudable by the expectations we’ve come to have of our government.

To be honest, though, it really shouldn’t be noteworthy that government lets us know how our money is spent – our taxes are one of the only things Americans don’t get a receipt for after all – and it’s definitely not appropriate to praise the transparency of new efforts like Recovery.gov if today is an example of how they implement. Right now Recovery.gov is basically useless to me. And as someone who really cares about how all this is going down, it’s quite frankly making me very angry.

And to be clear, the issue of bad data and extends far beyond our lovely ham examples.

What’s just as bad is that if you enter the contract number for the $1.2 ham in question into http://usaspending.gov, you get an incomplete record of Clougherty Packing and the contract as a whole.  That is, unless Clougherty’s annual revenue is in fact $1. (Hint: It’s not.)

And these hams are just one case.   Check out this contract for $5.8M to Keplin Construction Service Inc that has no description at all.

picture-7It’s just so completely unacceptable that I’m baffled at how it happens in the first place.

Recovery.gov already has issues for being far behind privately run Recovery.org in reliably tracking up-to-date stimulus spending. If .gov doubles down with inaccurate, unforgivably-incomplete data that requires a press release to clarify each entry …well then, why the heck do we even bother?

I’ve been asked by friends, “well, what do you expect? It’s the government…”

And you know, I consistently respond with, “yes it is, and I have confidence they have it in them to do a good job.” I still believe it’s true. There are a lot of smart folks working in agencies across the government right now and, to be fair, most haven’t been there that long as the new administration has hired up.  I know, specifically, that there are smart people at the USDA and at the Department of Interior.

Sadly however, the fact that there are quantifiably smart people in government is mostly serving to make issues like these examples on Recovery.gov even worse. It’s much more frustrating and disappointing as a result.

So here’s what I want …and what I think I should be able to expect.

I want to feel confident that when I get excited because my government announces they are committed to a new level of transparency and citizen engagement through initiatives like recovery.gov, I can trust that their commitment is true – and that it will actually matter. I want to be able to trust that my government’s leaders are going to invest in technology wisely and responsibly, and build useful tools that reliably provide me with the information I want or need — the information that I’m paying for.

I hope the folks down the street in the White House and in every agency can start getting their act together fast.  They’ve built a lot of good will by announcing and launching new online tools like Recovery.gov, but they’re losing it quickly by allowing it to be populated with crap.

Categorized in:
Share This:
  • kathy

    As a former fed, I would like to point out that this is the first time this information has been brought together in one place. Most of the information that feds have to enter on a purchase are not used for anything the fed needs. However, they have to enter it into a database because they have to. This often is the equivalent of the 10 times you have to fill out the same form for your medical appt. Once the systems begin to talk to each other, the data will improve. Each participant who needs data will enter what is useful to their program and the value of the data will drive the quality. Until that happens [which is a slow and painful process] you have to depend on the care of the multiple people entering data they don’t need for a purpose they don’t have. A friend used to say that whatever is not worth doing is not worth doing well. While public knowledge is a very good goal and an appropriate purpose, most people are more concerned with trying to do the multiple tasks that are piled onto them everyday, many of which seem to the person as make work because they don’t use the info they are providing. Also, the feds have almost done away with clerks so the information is often entered by a budget or contract analyst that has much more than data entry on their plate. Give this time, it will get better as the IT becomes more integrated and useful.

  • This issue has garnered a lot of attention, which is badly needed when it comes to talking about federal spending data quality. Unfortunately, this is far down the list of data quality issues. Since we created FedSpending.org in 2006, we’ve discovered the SBA publishes loans data for people who never received loans, the USDA was publishing citizen SS#’s in their data for decades without even knowing it, contracts and grants are listed in congressional districts that do not match the state listed for the spending, and many other, more serious data problems.

    I agree this type of error with the ham should be fixed. And I hope the increased attention this is bringing to data quality problems in the federal government spending data will help tackle some of the more major issues.

  • Cynthia

    Try and make an error like this when filing your tax returns. Oops! I’m quite certain they will not accept a letter of apology. They will want fees and penalties collected against your misjudgment. The government should be accountable for accurate information just as they hold us accountable.

  • SSO

    @Kyl,

    Here’s one way to solve the problem:

    Don’t buy the ham in the first place.

    My tax dollars not spent … no need to hire two people to check the data. Problem solved!

    @Jake,

    I wonder, why doesn’t Recovery.gov use NSNs (National/NATO Stock Numbers)? Surely, when the federal government buys 300,000+ hams, they have an NSN or other number assigned to the item to track the inventory/fulfillment.

  • Lori Collins

    Appreciate you bringing this into focus, Jake. And for the follow-on dialogue it triggers. The conversation gives me hope that people are paying attention and the spirit of accountability is ultimately there, albeit flawed.

  • So if bad data are the problem, what is the solution? Pay a person to double check every point of data on the site?

    That’s not terribly practical… What if they miss an error? We’re right back where we started. Except now we’re also out one federal employee’s salary. (Of course the answer here is: Hire two, and have them double check each other’s work for a mere 200% of the original cost!)

    I couldn’t agree more that this information needs to be accurate. I think where we disagree is where to draw the line. I’m pretty sure we’re both on the same page that bad data can be worse than no data whatsoever, as bad data begets bad intelligence, which in turn begets bad decisions. Moreover, the fact that they haven’t changed the record on the Ham page to be more clear is, to me, a blatant misstep. That should have been “fixed” — or more accurately, “clarified” — before any press release was considered.

    All that being said, my argument boils down to me thinking that it’ll take a long time (more than ten years) get every record 100% right 100% of the time. With so many data points, and so many moving pieces, it’s a very difficult, if not potentially impossible task.

    If we were to wait until everything was perfect, there’d never be a Recovery.gov at all.

    I just have one last thing to say, and it doesn’t really fit in anywhere, so I’m tacking this on at the end. In his speech at Personal Democracy Forum, Vivek Kundra (the nation’s Chief Information Officer) was talking about the launch of some of these transparency tools. I’m probably recalling his words incorrectly, but he said, basically, that when the people responsible for maintaining some of this information were told that it was going to be made public on the website, these reports and stats — which, up to then, were assembled in a perfunctory fashion — started to improve, because their authors knew that they’d face public scrutiny.

    Perhaps this is the point I’m driving at: Your logic leads us to a chicken or the egg scenario. If we don’t make the data public, the people producing them have little incentive to ensure accuracy. And if the data aren’t accurate, then the people producing them won’t want to make them public.

    We’re seven months into a dramatic shift in how the US Government views transparency and technology, if these problems are around in two years, it might be time to talk. But for now, I say let the sunlight keep shining, keep the pressure on to fix problems, and keep rebuilding the site, because it can always be better.

  • Jake Brewer

    Thanks for your comments, guys. Kyle, I do agree that it’s better for this info to be online than for it to be stuck in a drawer somewhere. That said, I don’t think that excuses these errors. If it were just one, I could overlook it, but it’s no where close to just one. If, as you say, “allowing people like you and me to point out all the places where there are disconnects between record and reality” is the point of these sites, then here we are doing just that. Thanks for joining me in that effort.

    Joe, to your point that most of the media focused on the ‘price of sliced ham’ so to speak: I think you’re right that they over reported a small point, and adding to it, they missed the big point entirely.

    The point here has absolutely nothing to do with ham, slicing or agriculture in general. It’s about accurate data, real transparency, and how we use it to hold gov’t accountable.

    The data our government is giving us in the one place they’ve given us to hold them accountable for stimulus spending is broken. The onus should be on THEM to tell us how they spent $1.2M (or $5.8M if you read the other example), not on us to interpret what “2 POUND SLICED” actually means.

  • JoeCitizen

    “Read as stated, this clearly leads one to believe that the U.S. government spent nearly $1.2M for 2 pounds of ham …sliced.”

    Yeah, but only if you are fricken retarded.
    Which, apparantly, most of the rightwing media are. Whats your excuse?

    Are you not embarrassed to put your name on such drivel?

  • We’re in a transitional phase here, where there are bound to be errors like this that will need to be corrected. I think in this case, you’re turning perfect into the enemy of good.

    With an organization as large as the federal government there’s no way we’ll have perfect data. It just takes one person’s slip up typing in a record, one miscommunication, and then the data are wrong.

    So rather than viewing this as a failure of Recovery.gov, I think it’s a great example of one of its usecases.

    If this information were confined to some drawer somewhere, it would still be labeled as $1 million for 2lbs of ham. By making it available — even in a non-perfect state — we’re allowing people like you and me to point out all the places where there are disconnects between record and reality.

    We need to “launch early, launch often”, to borrow a phrase from the startup world, and understand that we’re going to be working out kinks, ferreting out bad data, and finding places where the effort can be improved. And at the end of the day, that improvement is part of the reason the site exists at all.