Trade in Goods and services

1
Trade in Goods and services

• A report of differences in two OECD datasets
• (Balance of payments and Annual National
  Accounts)

OECD statistics directorate
2
Introduction

• Eurostat/European Central bank Task force has
  investigated conceptual and actual differences
  between b.o.p statistics and the Rest of the
  World account.
• Generally less understanding of differences
  between b.o.p. and national accounts series than
  goods in b.o.p. and Foreign trade.

3
The OECD comparison

• 10 annual series in national currency and current
  price were extracted from OECD MEI database and
  ANA database. (mid may 2007)
• Examination of the actual differences between the
  series in SNA and the series in BOP datasets
  (i.e. series x (SNA) in year t series x (BOP))
  in year t
• Series compared Goods, Services, Income,
  Transfers, Capital.
• Aim Investigate if possible to reproduce the
  conclusions drawn by Eurostat noting some EU
  peculiarities and adding some OECD specific
  remarks.

4
Preliminary remarks

• Most conceptual differences are expected to
  disappear after the current update on the
  international statistical manuals (1993 SNA and
  BPM5)
• Specific compilation practices have impact on
  consistency (including revision practices)

5
Main conclusions 1/2

• For most countries, trade in goods series have
  the best match in relative terms where most
  national mean differences were less than one
  percent between SNA and BOP.
• The second best match was for trade in services
  where most national mean differences were less
  than three percent.
• The third best match was for income where most
  national mean differences were less than ten
  percent.

6
Main conclusions 2/2

• Current transfers and capital account usually
  poorly match.

7
TF conclusions regarding differences in trade in
goods

• Transit trade The TF advises to register the gap
  between imports and export in some entities
  involved in transit trade to be recognized
  preferably as a service in b.o.p. statistics.
• Repairs Repairs should be recorded net both in
  b.o.p. and in the SNA.
• Reclassifications between goods and services
  This relates to certain types of goods recorded
  in external trade statistics which are
  reclassified to service in RoW by some countries
  while they are retained in goods in b.o.p.

8
Exports and imports of Goods Mean Differences
SNA-BOP as of BOP
9
OECD remarks from summary table on goods
comparison 1/2

• For exports of goods Zero or negligible mean
  differences 1996-2005 for Canada, Czech
  Republic, Finland, Iceland, Japan and UK. The
  largest mean differences were for Greece
  25.5, Turkey 13.2, and Portugal 5.8. The
  remaining countries had mean differences of
  below 4 with a slight majority having a positive
  sign.
• For Imports of goods, there was zero or
  negligible mean differences for Canada,
  Iceland, United Kingdom. The largest mean
  differences are found for Greece 17.7, Ireland
  4.87, Australia 3.85 and Turkey 3.42

10
OECD remarks from summary table on goods
comparison 2/2

• Greece The larger differences appearing may be
  explained by recent OECD SNA revision that has
  not yet been reflected in the OECD b.o.p.
• Regarding Turkey, the difference due to the
  country following SNA 68 for NA data.
• The differences shown for the Netherlands, (1.92
  for credits 0.94 for debits) can be according
  to the TF paper related to transit trade.

11
TF conclusions regarding differences in trade in
services

• Incidence of methodological differences higher
  than for goods. Differences are principally
  linked to the treatment of construction services
  and FISIM.
• On Construction The TF noted on Construction
  that BOP is in line with SNA 93 not with ESA 95.
• On FISIM At present FISIM is identified in
  national accounts, as recommended by ESA 95 and
  SNA 93, and not in BOP. Therefore a
  methodological difference exists and FISIM are
  classified under (financial) services in the RoW
  account, while they are indistinctly recorded
  under investment income in BOP.

12
Exports and imports of Services Mean
Differences SNA-BOP as of BOP
13
OECD remarks from summary table on services
comparison

• Expected B.o.p. services figures to be smaller
  than NA figures (non inclusion of FISIM). The
  case for 12 countries, otherwise not clear.
• So no obvious link can be drawn from the absence
  of FISIM in b.o.p. service data and the relative
  size of the figures compared to SNA figures.
• Zero or negligible mean differences for the
  United Kingdom which has totally integrated the
  b.o.p. and SNA processes. FISIM are also already
  integrated into b.o.p. data.
• For Exports, the largest mean differences
  appear for Japan -27.98, Portugal -22.88,
  Austria, -15.82 Luxembourg 11.34 and the
  Netherlands11.07.
• For imports, they were shown for Austria -36.06,
  Japan -16.54, the Netherlands 13.84,
  Luxembourg 13.80.

14
TF conclusions regarding differences in trade in
income

• Regarding Income from collective investment
  institutions (CIIS), the TF notes that the main
  methodological difference between BPM5/SNA93 and
  ESA 95 is due to the treatment of undistributed
  earnings. While ESA 95 requires the recording of
  distributed and retained earnings, BPM5 and SNA
  93 only provide for the recording of distributed
  earnings.

15
OECD remarks from summary table on income
comparison

• No striking influence of FISIM on the income data
• zero or negligible mean differences for United
  Kingdom (for exports). Norway also shows a
  perfect match since 1999.
• For Exports, the largest mean differences were
  for Turkey 65.82, the Netherlands 52.99,
  Slovakia 46.36 and Canada -30.50.
• For Imports, they were shown for the Netherlands
  53.41, Greece 26.52, Japan 23.82 and Canada
  -17.49.

16
Exports and imports of Income Mean Differences
SNA-BOP as of BOP
17
TF conclusions regarding differences in transfers

• Taxes on income and wealth The treatment of tax
  refunds differs between BPM5 and ESA 95.
• BPM5 states that current taxes on income and
  wealth should be recorded on the credit side,
  whereas refunds of taxes are included in the
  debit side.
• ESA 95 states that the total value of taxes
  should be reduced by the amount of any rebates or
  refund (i.e. net recording).

18
OECD remarks from summary table on transfers
comparison

• The amounts are much smaller than for the 3 other
  items. However the differences can still be very
  large
• There was zero or negligible mean differences
  for Canada only. Relatively close match for
  Korea, Mexico, Norway and Poland.
• For Exports, the largest mean differences were
  for Greece -98.37 and Spain -79.69
• For Imports, they were shown for Spain -44.46
  and Switzerland 41.34.

19
Exports and imports of Capital Mean Differences
SNA-BOP as of BOP
20
OECD Remarks on Capital series

• Data poorly correspond.
• Zero or negligible mean differences for Canada.
• The largest differences in exports were shown for
  Slovakia -449 and Sweden -127.57 and for
  imports were shown for the Czech republic
  -326.06 and Ireland -106.87

21
(No Transcript)
22
Questions?

• Delegates are asked if such comparisons are
  deemed useful and if more of them should be done
  (on which series ) ?
• Also, should the same comparison be done in say 2
  years in order to assess if the discrepancies
  between National Accounts and balance of payments
  series are growing or diminishing ?