Analyzing Services Trade and Policy: Overcoming Limitations of the Data Aaditya Mattoo WTO Data Day, 18 May 2009.

Analyzing Services Trade and Policy:

Overcoming Limitations of the Data

Aaditya Mattoo

WTO Data Day, 18 May 2009

Four broad questions

• What is the pattern of trade in services and how is it determined?

• What are the barriers to trade and how big are the gains from eliminating them?

• What are the elements of successful reform of services trade policy, and how do they differ across sector and country?

• What should be the focus of international cooperation on services?

I. What is the pattern of trade in services and how is it determined?

International Transactions in Services and Existing Statistical DomainsMode of Supply Relevant Data Source

Cross border supply BoP service statistics (categories other than travel)

Consumption abroad BoP Statistics (mainly the travel category)

Commercial presence Production, FDI and FAT statistics

Presence of natural persons

BoP and FAT Statistics

BoP data: Shared interest in cross-border trade in business services

Regional distribution of business services exports, 1990-2005

Average growth rates of business services exports, 1995-2005

Regional Distribution of Business Services Exports

223 7 9

256

79

4 11 8

380

86

10 17 12

530

171

3523 23

0

100

200

300

400

500

600

East Asia & Pacif ic South Asia Latin America & Caribbean Middle East & Africa OECDs

(in b

illio

n of

US

dolla

r)

1990 1995 2000 2005

906 Average Growth Rate of Business Service Exports for Selected Countries during 1995-2005

9.3

9.4

9.7

9.8

10.6

10.6

10.9

10.9

11.5

11.9

12.2

12.2

12.3

13.6

14.4

14.7

15.1

16.6

25.4

31.6

0 5 10 15 20 25 30 35

United States

Norway

Sweden

Mauritius

Malaysia

Morocco

New Zealand

Nigeria

Jamaica

United Kingdom

Spain

Finland

Malta

Israel

Argentina

Romania

Brazil

China

India

Ireland

(%)

OECD dominates… ..but developing are among the most dynamic exporters

What explains services dynamism?National data show human capital matters:

Evidence across Indian statesServices output and Education: averages over 1980-00

OR

AP

BH

HY KNKR

TN

UPWB

GJ

MPRJ

PJ

MH

0

1000

2000

3000

4000

5000

6000

0 0.001 0.002 0.003 0.004 0.005

Tertiary educated per cpaita

Ser

vic

es

ou

tpu

t p

er c

ap

ita

Source: Amin and Mattoo (2006).

What explains services dynamism?National data show institutions matter:

Evidence across Indian statesT&D Loss and Services per capita: averages over 1980-00

BH

OR

TN

WB

PJ

KN

UP

MP

KR

GJ

AP

HY

RJ

MH

R2 = 0.45

0

1000

2000

3000

4000

5000

6000

10 15 20 25 30

T&D Losses (%)

Se

rvic

es

pe

r c

ap

ita

Source: Amin and Mattoo (2006).

Goods85%

Services15%

US Monthly Imports of Total Goods and Services, Shares in Total Exports (2008) and Year-on-Year Growth Rates

Source: Borchert and Mattoo (2009).

US high frequency trade data reveals the crisis-resilience of trade in “other private services”

Why is services trade more resilient?

• Demand for traded services is less cyclical: what does national data on output and employment show?

• Services trade and production are less dependent on external finance: what does firm or industry level data show?

• Little explicit protection, but changing political climate and widening boundaries of the state may introduce a national bias in firms’ procurement and location choices: what does policy or firm-level data show?

“It ain't what you don't know that gets you into trouble. It's what you

know for sure that just ain't so.”

• GDP data in poor countries and the secret of steady growth

• GDP data in rich countries and services productivity growth

• India’s services exports to the US and US imports from India

• Trade in telecom, banking and insurance: What are we really measuring?

II. What are the barriers to trade and how big are the gains from eliminating them?

Restrictiveness of Services Trade Policy, 2007-08

ARG

AUSAUTBEL

BRA

KHM

CANCHL

CHN

COLCZE

DNK

ECU

EGY

FIN

FRA

DEUGHA GRC

HUN

IND

IDN

IRL

ITA

JPN

JOR

KENKOR

LTU

MYS

MEX

MNG

MAR

NLD

NZL

NGA

PAK

PER

PHL

POL

PRT

RUS

SAU

SEN

ZAF

ESP

LKA

SWE

TZA

THA

TTO

TUN

UKR

GBRUSA

VEN

10

20

30

40

50

60

Re

stri

ctiv

en

ess

of serv

ices tra

de p

olic

ies

0 10 20 30 40GDP per capita (thousand)

GDP per capita,PPP (constant 2005 international $)

Source: Gootiiz, MattooSource: Gootiiz and Mattoo (2009).

Remedying the gaps in data on services trade policy

Step 1: Times Series Data on Policy: In recent years, India has reformed its services sectors and

significantly liberalized FDI

Newly created index of services reform

0

0.5

1

1.5

2

2.5

3

3.5

4

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

Banking Insurance Telecom Transport

Newly created index of services reform

0

0.5

1

1.5

2

2.5

3

3.5

4

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

Banking Insurance Telecom Transport

0

5

10

15

20

25

30

35

40 Growth Rate (CAGR)

36%

20%

FDI in Services*

FDI in Non-services

India: FDI in services sector is growing faster than in non-services . . .

(1992/93 values are indexed to 1; Unit: US$)

1992

/93

1993

/94

1994

/95

1995

/96

1996

/97

1997

/98

1998

/99

1999

/00

2000

/01

2001

/02

0

5

10

15

20

25

30

35

40 Growth Rate (CAGR)

36%

20%

FDI in Services*

FDI in Non-services

India: FDI in services sector is growing faster than in non-services . . .

(1992/93 values are indexed to 1; Unit: US$)

1992

/93

1993

/94

1994

/95

1995

/96

1996

/97

1997

/98

1998

/99

1999

/00

2000

/01

2001

/02

Step 2: Econometric analysis using firm-level data suggests services policy reform

benefited manufacturing industriesGains in Annual TFP Growth After Services Reform

0

1

2

3

4

industries withlimited

dependence on banking

industries with high

dependence on banking

industries with limited

dependence on telecomms

industries with high

dependence on telecomms

Percent Gains in Annual TFP Growth After Services Reform

0

1

2

3

4

industries withlimited

dependence on banking

industries with high

dependence on banking

industries with limited

dependence on telecomms

industries with high

dependence on telecomms

Percent

New study based on:•panel data for 4,000 Indian firms for the 1990-2005 period, and •input-output linkages, finds that

banking, telecommunications and transport reforms all have significant positive effects on the productivity of manufacturing firms

Source: Arnold, Javorcik, Lipscomb and Mattoo (2008).

Welfare Gains from a 3% Increase in Developed Countries' Temporary Labor

Quota

0

20,000

40,000

60,000

80,000

100,000

Developed Res t of the world

Unskilled Labor Skilled Labor

Note: Data in million US$

Source: Walmsley and Winters (2002)

Using CGE Models to Produce ex ante Estimates of Gains from Mode 4 Liberalization

The paucity of data on performance

• The different dimensions of performance: productivity, prices, quality, variety, access

• The distributional consequences: across income groups and regions

• The costs of adjustment: for firms and individuals

III. What are the elements of successful reform of services trade policy?

Case studies using survey data: Openness has not improved access to financial

services in Zambia

• Banking was liberalized before establishing a proper regulatory framework. 1991-1994: Ten new bank licenses issued1995-2001: Nine bank failures, causing estimated losses

equivalent to 7 percent of GDP.

• Foreign banks today account for over two thirds of total assets, loans and deposits. But credit to the private sector is only 8 per cent of GDP -

lower than in 1990 and in most other Sub-Saharan African countries.

Only 5,000 people hold 90 percent of loans.

Source: Mattoo and Payton (2007)

The sequence of telecom reform in selected countries

Country 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Malaysia Privatization Competition Regulation India Privatization Competition Regulation Argentina Privatization Competition Regulation Brazil Privatization Competition Regulation El Salvador Privatization Competition Regulation Nigeria Privatization Competition Regulation South Africa Privatization Competition Regulation Uganda Privatization Competition Regulation

Source: World Bank/ITU Telecommunications Policy Database & Fink, Mattoo, Rathindran (2001), forthcoming. Note: Competition refers to the fixed-line local services segment.

Step 1: Compiling cross-country data on the sequence of policy reform: Privatization, competition and regulation in basic telecommunications

Source: World Bank/ITU Telecommunications Policy Database & Fink, Mattoo, Rathindran (2001), forthcoming.

Growth in mainlines

12.40%

25.50%

0.00%

5.00%

10.00%

15.00%

20.00%

25.00%

30.00%

Privatization before competition

Competition before privatization

Step 2: Combining policy data with ITU data for an econometric analysis of the impact of alternative sequences of reform:

Delaying competition after privatization can dampen growth of telecommunications

Understanding the politics behind policy

• Zambia telecommunications and transport

• US financial regulation and maritime transport

Two examples of contrasting transparency:

Using data to measure the exercise of political influence in the US

Campaign Contributions during the 2008 Election Cycle by US Commercial Banks

Rank Organization Amount Democrats Republicans 1 JPMorgan Chase

& Co $4,280,866 61% 39%

2 Citigroup Inc $3,467,047 65% 35% 3 American

Bankers Assn $3,067,061 43% 57%

4 Bank of America $2,200,793 56% 44% 5 Wells Fargo $1,578,597 50% 50% 6 Deutsche Bank

AG $1,356,754 72% 28%

7 Independent Community Bankers of America

$1,013,075 53% 47%

Total contributions by all US commercial banks

$36,329,595

IV. What should be the focus of international cooperation on services?

Transport and Mode 4 are heavily protected but neither is being seriously negotiated

Restrictiveness index: by income group

0.00

10.00

20.00

30.00

40.00

50.00

60.00

70.00

80.00

90.00

Lowincome

Low ermiddle

UpperMiddle

Highincome

Fin. Serv

Telecom

Retail

Transport

Professional

Source: Gootiiz and Mattoo (2009).

Remedying the gaps in data on services trade policy by sector

Uruguay Round Commitments, Doha Offers and Actual Policy

ARG

AUS

BRA

KHM

CAN

CZR

CHL

CHN

COL

ECU

EGY

_EUN

GHA

HNG

IND

IDN

JPNJOR

KEN

KOR

LTN

MYS

MEX

MNG

MAR

NZL

NGA

PAK

PER

PHL

SAU

SEN

ZAF

LKA

TZA

THA

TTO

TUN

USA

VEN

ARG

AUS

BRA

KHM CANCZR

CHL

CHN

COLECUEGY

_EUN

GHA

HNG

INDIDN

JPN

JOR

KEN

KOR

LTN

MYS

MEX

MNGMAR

NZL

NGA

PAK

PER

PHL

SAU

SEN

ZAF

LKA

TZA

THA

TTOTUN

USA

VEN

ARG

AUS

BRA

KHMCANCZRCHL

CHN

COLECU

EGY

_EUNGHA

HNG

IND

IDN

JPN

JOR

KENKOR

LTN

MYS

MEX

MNG

MAR

NZL

NGA

PAK

PER

PHL

SAU

SENZAF

LKATZA

THA

TTO

TUN

UKRUSA

VEN

020

40

60

80

10

0

0 10000 20000 30000 40000GDP per capita in 2006 (constant 2005 USD)

Restrictiveness of GATS commitments Fitted values

Restrictiveness of Doha Offers Fitted values

Restrictiveness of actual policies Fitted values

Source: Gootiiz and Mattoo (2009).

Assessing the Value of Trade Agreements

Looking ahead: Cost-Benefit Analysis of Data Itself

How valuable from a policy-perspective are certain types of information and how much would it cost to collect the relevant data with a certain degree of reliability ?

• Policy makers and negotiators: What are the key policy questions? E.g. to have a better sense of exports of IT-enabled services from India; of tourist expenditures in the Caribbean; of the importance of foreign banks in Eastern Europe; of the sources of remittances for Bangladesh and Yemen? Where is the protection? Where are the gains from cooperation – regionally, multilaterally?

• Economists and analysts: What types of data can help address the specific questions?

• Statisticians: What is the cost of collecting specific types of statistics?