Attachment 2: Written comments from data users

Attachment 2
From: Lipsey, Robert [mailto:[email protected]]
Sent: Monday, April 23, 2007 1:44 PM
To: Galler, David
Subject: Bank survey forms
Dear Mr. Galler,
I have not worked much with bank FDI data or thought much about its problems, so you should
probably not give a lot of weight to my reactions.
From the little I know about this sector, my impression of the problems is that the most important
ones, stemming from the intangible nature of the production and trade, are those of identifying the
location of production and the direction of trade. For that reason, I am uncomfortable with the apparent
identification of “country of location” of the activities with the location of physical assets (Item 4). That
basis for location makes much more sense for nonbank affiliates in manufacturing and mining, where
inputs from physical capital are of great importance than for industries, such as banking, where physical
capital inputs are of very minor importance. The same issue arises for many service industries, where
skilled labor, rather than physical capital, defines the actual location of activity. My preference would be
to have a location question based on employment or payroll, which I think identify the location of a
service affiliate’s activity in a more justifiable way.
I see that there is no analogue in the bank form to the Part III of the nonbank form. I can’t think
of a good analogue, especially since the nonbank form is emphatic in its reliance on goods shipped across
US borders. It is not clear to me how data collected in these surveys, where location is defined in terms
of physical plant, will relate to BEA’s survey data on service trade.
Sincerely,

Robert E. Lipsey
Director, New York Office
National Bureau of Economic Research
365 Fifth Avenue, Suite 5318
New York, NY 10016-4309
Tel: (212) 817-7961 / 7955
Fax: (212) 817-1597
[email protected]

From: Vladimir Gololobov [mailto:[email protected]]
Sent: Friday, April 27, 2007 5:57 PM
To: Galler, David
Cc: John Goyer
Subject: Comments on BEA's Forms
Dear David,
Thank you for the opportunity to comment on the draft forms for bank US parents and affiliates, and bank
affiliates on non-bank parents. I have one general observation: BEA might consider collecting more
detailed information on sales by parents and affiliates, for example by providing sales by country. The
same goes for the proposed employment and investment surveys: it would be helpful if these surveys
break data by country where possible. These data will be especially helpful to us to advocate FTAs.
For your information, I’m forwarding you the attached CSI statement for this week’s Appropriations
hearing on the FY08 budget. We asked for sufficient funds for the Census and BEA, so I hope those
projects will be fully funded.
Thank you,
Vladimir

C. FRITZ FOLEY

|

ASSISTANT PROFESSOR

May 10, 2007
David H. Galler
Chief
Direct Investment Division
Bureau of Economic Analysis
Dear Mr. Galler,
I am writing to reply to three letters I have received from you recently, each asking for my views concerning
the design of BEA surveys. I have broken out my views into three parts below, each part responding to one of
your letters.

Proposed Changes to the BE-12 Survey
In your letter dated April 30, 2007, you asked me to provide you with my comments and suggestions
concerning the proposed changes to the benchmark survey of foreign direct investment in the U.S. I support
most of the proposed changes, but I have one major objection and a few suggestions.
My biggest objection to the proposed changes relates to the proposal to increase the reporting thresholds. The
increase in these thresholds that has occurred over the last ten to fifteen years has dramatically decreased the
amount of data firms are required to report and therefore threatens the value of the BEA data. Changes in the
requirements concerning which firms must report which data items has also contributed to this decline. I
therefore would suggest that the reporting requirements not be changed.
In 1992, all U.S. affiliates with sales, assets or net income with an absolute value in excess of $1 million filled
a complete data form. The proposed changes to the survey would raise this threshold to $40 million. Proposed
requirements for reporting data on the long form data are also far too high. I believe that the long form was
only introduced within the last fifteen years—prior to that all companies effectively filed all the information
captured on the long form. Since its introduction, thresholds have grown in a way that has significantly
reduced the sample of firms that report details of their financial and operating activity.
These changes have the potential to destroy the value and accuracy of many aspects of the BEA data. While
BEA may be able to claim that, on a value weighted basis, a large fraction of multinational activity is studied
in detail, these changes undermine our understanding of activity in many countries, industries, and types of
firms. Multinationals in smaller countries and in industries without significant economies of scale will not be
covered well by the BEA data. Certain data items in country/industry cells could also be misleading if there
are lots of small affiliates that do not report the item.
Under the proposed changes, it is only possible to track a few data items through time for smaller affiliates,
thus limiting the value of research using the firm level data. Researchers will not be able to perform detailed

SOLDIERS FIELD

|

BOSTON, MA 02163

|

Ph 617.495.6375

|

Fx 617.495.7659

|

[email protected]

|

G E O R G E F. B AK E R F O U N D AT I O N

studies of how changes in the international economic environment and policy changes affect firm operating
and financial decisions using large samples if reporting thresholds continue to increase.
It does not make sense to me that BEA should reduce coverage of multinational firms at the same time that
information on multinationals is of increasing importance to academics, policy makers, and business leaders.
If it has been difficult for BEA to obtain the resources needed to maintain the accuracy and completeness of the
data, I wonder if BEA data users could help decision makers understand the value of these data. I am happy to
discuss this possibility with you.
I have some other suggestions concerning the proposed changes as well. I would not double the threshold for
reporting merchandise trade on the BE-12(LF) form. Although these data have not been widely used by
researchers, they are unique, and I believe that researchers will use them extensively in the future. I would
suggest leaving this threshold at $500 thousand.
I would suggest collecting information on capital expenditures in Part I of the short form and on the BE-12
Mini form. Offshoring issues relate not just to where employment is located, but also to where capital is
invested, so capturing this item for smaller affiliates is crucial.
I would also suggest collecting information on dividends in Part I of the short form and on the BE-12 Mini
form. Dividend repatriations are a significant flow of funds out of and into the U.S. and have been the subject
of considerable academic and policy interest over the last two decades. For example, the American Job
Creation Act of 2004 temporarily reduced the taxes the U.S. charges on earnings repatriated to the U.S. As
debates about how multinational firms should be taxed continue, it will be useful to have information on the
dividends paid by U.S. affiliates to inform the debate.

Shipped versus Charged Basis for Trade Data
In your letter dated May 3, 2007, you explain that there is a proposal to report trade data on a charged as
opposed to a shipped basis. I do not support this proposal, and I would advise you to continue to collect and
report trade data on a shipped basis. I do appreciate that tracking and providing data on a shipped basis is
costly for firms and BEA. I think this argument is the strongest one for changing reporting. However, I
believe that BEA’s trade data are of more value to current research efforts in international economics if they
are reported on a shipped as opposed to charged basis. Let me explain my view on this.
Studies of transfer pricing that make use of the BEA data tend to focus on how foreign taxes affect levels of
reported income (see Hines and Rice (1994) or Desai, Foley and Hines (2004a)), levels of investment and
economic activity, and the use of different factors of production(see Desai, Foley and Hines (2004b). It would
be helpful to have trade data on a charged basis if data on a shipped basis were also collected, or if trade data
provided details not just on values but on prices and quantities. This kind of data would allow a researcher to
see if firms shift profits to low tax locations. I do not believe that simply knowing where traded goods are
charged will allow one to make much headway. Asking for additional details for trade data would create
additional burden for reporters, so I do not think it is feasible.
Studies concerning how ownership patterns affect trade and how endowments affect trade rely on having trade
data that is reported on a shipped basis. For example, consider Antras (2005). Antras (2005) empirically
documents and theoretically explains patterns in the share of intrafirm trade by industry. For his study, Antras
must be able to compare levels of intrafirm trade with levels of total trade. He uses the BEA data and relies on
the fact that these data and the Census data are both reported on a shipped basis. The literature on the impact
of factor endowments on trade also relies on knowing where traded goods are going to and coming from as
opposed to who is paying for them.
In short, I believe that trade data are more valuable to researchers if they are collected on a shipped as opposed
to charged basis. If reporting burdens are a major concern, I wonder if it would be possible to try to coordinate

data collection efforts with the U.S. Census Bureau and the U.S. Customs Bureau. If it were possible to
coordinate data collection efforts, reporting burdens for companies might fall and the accuracy of the data
might improve.

Covering Banks in the Annual BE-11 Survey
In your letter dated April 12, 2007, you indicated that BEA intends to cover bank U.S. parents and their bank
and nonbank affiliates in future Annual Surveys of U.S. Direct Investment Abroad. I think this is a great idea.
I have not seen much academic research that studies the BEA data on multinational banks, and this may in part
be a consequence of the fact that data are only available for certain years. I do have a few suggestions on the
reporting requirements and the data collection form.
For nonbank affiliates, I wonder why BEA would not ask bank parents to file the same BE-11A, BE-11B, and
BE-11C forms subject to the same reporting requirements as other multinational parents. It seems to me that
an affiliate in a manufacturing industry with a parent in banking should file the same information as any other
affiliate in manufacturing.
If nonbank affiliates filed these forms, this would provide more flexibility to design a BE-11B(BNB) annual
survey form specifically for bank affiliates. In the proposed BE-11(BNB) form, I am not sure how a bank
would fill out item number 40 which captures information on sales by location since deposit taking entities do
not really have sales. For nonbank affiliates, the $500 million reporting threshold also seems so high that very
few of these affiliates will report any data.
Thinking about banking data raises another issue for me. My understanding is that private equity firms are
increasingly making international investments that would trigger their being classified as multinational parents.
I do not think they would be classified in industry 5221, so I do not think these kinds of firms are affected by
the proposed changes concerning bank parents. However, I hope that BEA is capturing their international
activity on an annual basis.
Finally, in order to address reporting burdens and to improve the accuracy of the data, I would encourage BEA
to try to coordinate their data collection efforts with the efforts of the Federal Reserve Board. I do not know
the details of what the Federal Reserve Board does in the sphere of international banking, but I do know that
they report data on the foreign activities of U.S. banks, and I wonder if there would be gains from working
more closely with officials based there.

In conclusion, thank you for asking my views on these issues. I have a deep respect for all of the hard work
that you and others at BEA do to collect and analyze data about our economy. I appreciate the support that
BEA has given me in my research, and I hope that you find my suggestions and comments to be helpful.
Sincerely,
C. Fritz Foley

References
P. Antras, “Firms, Contracts, and Trade Structure,” Quarterly Journal of Economics (2003)y 1375-1418.
M.A. Desai, C. F. Foley and J. R. Hines Jr., “The Costs of Shared Ownership: Evidence From International
Joint Ventures,” Journal of Financial Economics (2004a) 323-374.

M.A. Desai, C. F. Foley and J. R. Hines Jr., “Foreign Direct Investment in a World of Multiple Taxes,”
Journal of Public Economics (2004b) 2727-2744.

J. R. Hines Jr., and E. M. Rice, “Fiscal paradise: Foreign tax havens and American business,” Quarterly
Journal of Economics (1994) 109:149-182.

From: J. Bradford Jensen [mailto:[email protected]]
Sent: Friday, May 11, 2007 11:57 AM
To: Galler, David
Subject: RE: Trade Data for U.S. Multinational Companies/BE-11 Survey/BE-12 Survey

Mr. Galler
I write in response to your letters of April 12 and April 30 and your
e-mail of May 3. Thank you for the opportunity to comment on the survey
changes being considered by the Bureau of Economic Analysis (BEA).
My interests in the data collected by BEA are primarily in the area of
trade, and trade in services in particular, so my remarks will focus on
those aspects of the surveys you asked me to comment on.

May 3 e-mail regarding change from "shipped" to "charged" basis:
I disagree with the proposed change. I think BEA should continue to
collect data on a "shipped" basis rather than a charged basis. In my
opinion, the strength of the statistical system is in collecting real
(as opposed to financial) information. I think the statistical system
should emphasize the collection of real side information. To this end,
maintaining comparability with the Census Bureau's data on imports and
exports seems important. I am also sympathetic to the argument that the
"shipped" basis information is more useful to researchers interested in
the impact of trade.
To reiterate, I disagree with the proposed change.

April 12 letter regarding proposed changes to BE-11 survey:
Extending coverage of the BE-11 survey to include bank parents and bank
affiliates of nonbank parents seems like an important extension.
Collecting data for the U.S. Reporter on the same BE-11A form as filed
for nonbank U.S. Reporters seems like the right approach. The
specialized form for bank affiliates also seems appropriate.
I would suggest one addition to the BE-11B(BNB) form. On the BE-11B (LF
and SF) Part III questions 98 and 99 ask about goods exports to and
goods imports from the U.S.
I would suggest adding a question to the BE-11B(BNB) form (and the
BE-11Bs too) that would be of the same format as question 98/99 on the
BE-11B but would ask about imports and exports of services.
It is not clear from the BE-120 form available on the BEA website who
the intended respondents are for the BE-120. It seems that it will only
be sent to parents or foreign owned affiliates in the U.S. I think

collecting information at the foreign affiliate level on imports and
exports of services to the U.S. is important. Will the BE-120 form be
sent to foreign affiliates of US MNCs?
Adding a question on service imports and exports to the affiliate survey
would provide a useful check on the information provided on the BE-120
form.

April 30 letter regarding proposed changes to BE-12 survey:
Regarding Item II.2 in attachment A (drop from the BE-12 the questions
on services transactions and replace with BE-120, 125, 180), this seems
to make sense to me. Having both affiliated and unaffiliated trade
reported on a consistent basis is a welcome change to the cross-border
services transactions program.
It is not clear from the write-up what type of question will be added to
identify companies that engage in cross-border services transactions
(Item III.1). I would recommend something along the lines of what I
recommended above for the BE-11 forms. Add a question similar to Part II
question 141 on the BE-12(LF) form that asks about services imports to
and services exports from the U.S.

I hope these comments are helpful. If I can provide additional
information, please let me know.
Thank you again for the opportunity to comment.
Brad

J. Bradford Jensen
Deputy Director
Peterson Institute for International Economics
1750 Massachusetts Avenue, NW
Washington, DC 20036

From: Ronald Fecso [mailto:[email protected]]
Sent: Friday, May 11, 2007 5:20 PM
To: Galler, David
Subject: BE-11 revision
Thank you for providing GAO with an opportunity to comment on your proposed revisions to the
reporting requirements for the BE-11.
I've called attention to the proposals to various persons and groups within GAO.
As you know, GAO if often a user of data of this type and thus our consideration focused on any current
or anticipated engagements for which the changes might materially impact our work.
In this perspective, we have no objections to the proposed changes.
Thanks again for soliciting our input.
Best wishes for the remaining activities in carrying out the proposed changes.
Regards,

Ron Fecso
Chief Statistician
U.S. Government Accountability Office
room 6K17
441 G Street, NW
Washington, DC 20548
202-512-7791
[email protected]

From: [email protected]
Sent: Tuesday, May 15, 2007 10:22 AM
To: Galler, David
Cc: [email protected]
Subject: Comments on new FDIUS Survey forms
Dear Dave,
Karen Laney-Cummings has passed your requests for comments on to me, as the person in our office who
deals most directly with the BEA investment data. It’s a pleasure to hear from you after all this time, and
I hope you’re doing well.
The following comments apply to both your letter of April 12, 2007, seeking comments on the inclusion
of bank parents and their affiliates in the annual surveys, and your letter of April 30, 2007, seeking
comments on the 2002 benchmark survey.
Regarding the revisions to the BE-11 Annual Survey form, I am particularly pleased to see that you will
be including banks and their affiliates in your annual surveys. I agree that the lack of data on bank
parents and their affiliates has been a significant gap in existing BEA annual data. However, I would
question the high exemption level of $500 million for bank affiliates, compared to what appears to be an
exemption level of $150 million for non-bank affiliates on the BE-11 instructions enclosed with your
letter. I’m concerned that we would be missing important information for smaller institutions. Is it your
understanding that the higher exemption level for banks will lead to the same share of overall affiliates
reporting? Do you have plans to include bank and their affiliates in the FDIUS annual surveys as well?
Regarding the benchmark survey, I’m quite pleased to see that you will be adding additional information
on services transactions and R&D to the surveys. We make extensive use of this information, and we’re
looking forward to the additional detail. I am particularly pleased to see the additional detail you will be
requesting on the activities of banks and bank holding companies, which seem likely to yield significantly
greater insight into the these firms’ activities.
I am concerned that by raising the exemption criteria, we will be losing valuable information about
smaller affiliates. While I understand that there is an issue of the reporting burden on small companies,
these companies can be quite an important and dynamic segment of the U.S. economy, and by exempting
these affiliates, it is not possible to tell what is missing from the data.
I don’t have any serious objections to the other items that will be dropped from the forms.
We appreciate the opportunity to comment on the changes to your survey forms. I look forward to
continuing to work with BEA in the future.
Best regards,
Laura
Laura S. Bloodgood, Ph.D.
Sr. International Trade Analyst for Investment
U.S. International Trade Commission
(202) 708-4726