The Corner

The Fight for Ex-Im Transparency: Citizens 1, Government 0

On Wednesday, I mentioned that in response to the removal of all the Ex-Im datasets from from Data.gov, Mercatus’s Andrea Castillo, Eli Dourado, Chris Koopman, and I have put all the data back online over at ExImUncensored.com for everyone to use. Our site includes the main dataset that was removed from Data.gov back in the summer, plus the state-by-state data that we got from Ex-Im.gov in a more user-friendly form.

As I wrote on Wednesday, I was hoping that it would entice the Ex-Im Bank to restore the dataset that was removed, and put up the 2014 numbers — perhaps in more complete, better form.

Well, today, two days after we posted the data online, I got two of my wishes. 

The new dataset, which was uploaded on February 12, 2015, is available for download on Data.gov! Their user experience could use a little work — the page simply provides four links to unlabeled “webpages” that  lead to unusually formatted downloads. (You can click on the first “Web Page” to download a .csv file). But I am overjoyed that a dataset is once again available to the public. We have updated ExIm Uncensored accordingly.

But the dataset now on Data.gov is a little different from what we had.

I’m pleased to report that there are some positive changes: The new dataset lists all of the new FY 2014 transactions that have occurred from the time the original dataset was pulled down until now, which is valuable. The new dataset also contains new fields that list the city and state for each primary exporter, making it easier to use as I’d asked. It would have been nice to have had this data from the very beginning, but let’s give credit where credit is due.

But there are big problems with the data that require explanation:

Three critical fields were deleted from the new dataset: Primary Buyer, Primary Source of Repayment, and Primary Supplier.​ This is a big deal — the bank should restore these missing fields. (Reminder that we have that information at ExIm Uncensored for 2007 to 2013.)

Further, the dataset only contains approved applications from 2007 to 2014. The original dataset contained all applications — approved and declined — over the same time period. This isn’t a huge deal, but it’s not the same data.

There are also two transactions that were listed in the original dataset that were removed in the new one. They are:

1.Deal 09294827  0001 (2007), a $9.7 million in insurance for an “Unknown” project in South Africa. Primary Applicant: Via Capital Services. Primary Lender: Unknown. Primary Exporter: Unknown. Primary Supplier: Unknown. Primary Borrower, Buyer, and Source of Repayment: Bell Equipment Company, S.A.

2. Deal 09551967ST1001 (2013): an insurance deal for an “Other Miscellaneous Durable Goods Merchant Wholes” project in the United States. Primary Applicant: Ohm International Corporation. Primary Lender: Hsbc Usa, Inc. Primary Exporter: Ohm International Corporation. Primary Supplier: Unknown. Primary Borrower, Buyer, and Source of Repayment: Public Utilities Corporation. This project lists “$0.00” for the disbursement amount, which is confusing in itself. The Bank should work to make data documentation resources available so that the public can make sense of their records.

Why were these records deleted? It could have been an honest error, or something else. A representative from the bank should explain the discrepancy so that we will know what happened.

But it all calls into question the bank’s data quality and management practices in general. I’ve been encouraging the bank to clean up and clarify their datasets for months now – the records are riddled with misspellings, missing data, and plenty of records that are simply listed as “Unknown” or “Various Businesses.” The public deserves better. 

This is a small victory for transparency, but the new dataset’s failure to list the Primary Buyer, Primary Source of Repayment, or Primary Supplier for each transaction is highly problematic. This information is critical for the public to have a transparent picture of all of the bank’s activities, and it’s still on ExImUncensored.com.

If this dataset was created in response to the transparency efforts of Ex-Im Uncensored, it’s unclear why the decision was made to remove three critical fields that tell the public the buyers, financiers, and suppliers for each deal.

I’ll try to get answers and keep you updated.

Veronique de Rugy is a senior research fellow at the Mercatus Center at George Mason University.
Exit mobile version