Broadband Economic Impact in Brazil: Broadband Economic Impact in - - PowerPoint PPT Presentation

▶
broadband economic impact in brazil broadband economic
SMART_READER_LITE
LIVE PREVIEW

Broadband Economic Impact in Brazil: Broadband Economic Impact in - - PowerPoint PPT Presentation

Sep 05, 2022 28 likes •652 views

Broadband Economic Impact in Brazil: Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis a Simultaneous Equations Analysis Hildebrando Rodrigues Macedo, Hildebrando Rodrigues Macedo, Anatel, Brazil Anatel, Brazil

slide-1
SLIDE 1

Hildebrando Rodrigues Macedo,

Broadband Economic Impact in Brazil: Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis a Simultaneous Equations Analysis

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 1

Hildebrando Rodrigues Macedo,

Anatel, Anatel, Brazil Brazil

Alexandre Xavier Ywata de Carvalho, PhD

IPEA, IPEA, Brazil Brazil 19 November 2011 19 November 2011

slide-2
SLIDE 2

Agenda Agenda

  • Introduction , Motivation and

Introduction , Motivation and Objectives Objectives Objectives Objectives

  • References

References

  • Methodology: Model

Methodology: Model

  • Results and Analysis

Results and Analysis

  • Conclusion and Comments

Conclusion and Comments

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 2
  • Conclusion and Comments

Conclusion and Comments

slide-3
SLIDE 3
  • Main Objective:

Main Objective: Evaluate the broadband Evaluate the broadband economic impact in Brazil using official economic impact in Brazil using official statistics data from statistics data from Anatel Anatel (the Brazilian (the Brazilian

Main Objective and Results Main Objective and Results

statistics data from statistics data from Anatel Anatel (the Brazilian (the Brazilian telecom regulatory agency) telecom regulatory agency)

  • Main result:

Main result:

  • 1

1p.p. p.p. (percentage point) (percentage point) increase in increase in broadband penetration related to GDP broadband penetration related to GDP

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 3

broadband penetration related to GDP broadband penetration related to GDP growth between growth between 0.038 to 0.18p.p. 0.038 to 0.18p.p., and , and GDP per capita growth between GDP per capita growth between 0.196 0.196 to 0.362p.p. to 0.362p.p.. .

slide-4
SLIDE 4
  • Too high economic impact found.

Too high economic impact found.

  • Data used are

Data used are “problematic” “problematic”

  • The

The lack of reliable data lack of reliable data for some of for some of

Limitations of the Study Limitations of the Study

  • The

The lack of reliable data lack of reliable data for some of for some of the variables and the need to the variables and the need to estimate the missing data estimate the missing data brings brings additional imprecision to the results, additional imprecision to the results,

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 4

additional imprecision to the results, additional imprecision to the results, which must be interpreted in a more which must be interpreted in a more cautious cautious way. way.

slide-5
SLIDE 5

Broadband Price Estimation Broadband Price Estimation

  • Model based on results

Model based on results of

  • f CETIC

CETIC survey on survey on maximum declared maximum declared price price for for internet adoption internet adoption. .

  • !

" # $ %

  • '!

! ()*%%+**+$* ,(*))%"

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 5
  • +
  • $

+ +$

  • $

"

  • !

,.!*

  • &
slide-6
SLIDE 6

"$

  • +#

+% $ "

  • 1/0
  • /

Broadband in Brazil: Overview Broadband in Brazil: Overview

$% +$ + + "$

  • #

%

  • +

+ + +$

  • $
  • !
  • /
/0 Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 6

%% +)$ ) #+# $%

  • #
  • $
  • +
  • "

# $ %

  • )

+

/ /

2

slide-7
SLIDE 7

Broadband Penetration Worldwide(2008) Broadband Penetration Worldwide(2008)

Country Country Broadband Penetration Broadband Penetration (number (number of accesses

  • f accesses

per 100 inhabitants per 100 inhabitants) ) Region Region Broadband Penetration Broadband Penetration in the Region (number in the Region (number

  • f accesses per 100
  • f accesses per 100

inhabitants inhabitants) ) Argentina Argentina 7.9 7.9

Brazil Brazil 5.91 5.91

Latin Latin America America 5.5 5.5

Brazil Brazil 5.91 5.91

Chile Chile 8.4 8.4 Colombia Colombia 4.2 4.2 Mexico Mexico 7.1 7.1 Canada Canada 29.0 29.0 North North America America 27.8 27.8 USA USA 26.7 26.7 Spain Spain 20.8 20.8 France France 28.0 28.0

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 7

Europe Europe 24.8 24.8 France France 28.0 28.0 Portugal Portugal 16.0 16.0 United Kingdom United Kingdom 28.5 28.5 Australia Australia 25.4 25.4 Asia and Asia and Oceania Oceania 14.0 14.0 South Korea South Korea 32.0 32.0 Malaysia Malaysia 4.6 4.6 South Africa South Africa 0.8 0.8 Africa Africa 1.6 1.6 Morocco Morocco 1.5 1.5

slide-8
SLIDE 8
  • Broadband Economic Impact in Brazil: a
Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 8

!"# $ %&

slide-9
SLIDE 9

' ' ! ! ( (

' ' ) )

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 9

*+' *+'

,-.# ,-.+- ,/01
slide-10
SLIDE 10

$*#+

$ %

  • ' !
  • 3

#4!*

  • 53*

+*#) ++*" *+ * +$*)" *+ *) $*$) )*) +#*) )*+%

+

  • "

# $ ' ! ' !

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 10

*# *") +*)+ *+) +* *+ * *+ *) $*$) "*+

  • +
6+ * 7+ 8 9* 7 " * 7" 8# * 7# $ * 7$ + * 7+
  • *
7 $ * :$* , !" 2 0
slide-11
SLIDE 11 $*#+ #$*) $ %
  • '!
  • 3

#4!* *

#*) +*# +$*)" * *+ #$*) +
  • "
# '! '! Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 11 *# *") +*)+ *+) +* $*" $*$$ *# *#% +*%$ *%% *+
  • +
6+ * 7+
  • *
7 8" * 7" # * 7# $ * 7$ + * 7+
  • *
7 $ * :$ * ,
  • 3 .4$ 2
slide-12
SLIDE 12
  • Koutroumpis

Koutroumpis (2009) (2009)

  • Broadband economic impact on

Broadband economic impact on 22 22 OECD countries OECD countries from 2002 to 2007. from 2002 to 2007.

  • Used

Used supply supply and and demand demand simultaneous equations simultaneous equations system system

Main Reference Main Reference

simultaneous equations simultaneous equations system system with endogenous variables to with endogenous variables to capture a capture a simultaneous effect simultaneous effect: :

  • Investments in telecom

Investments in telecom lead to lead to economic growth , economic growth , at the same at the same time time economic growth economic growth demands demands

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 12

time time economic growth economic growth demands demands telecom investment. telecom investment.

  • Result: approx.

Result: approx. 0,025p.p 0,025p.p. de . de GDP GDP increase increase for each for each 1p.p. 1p.p. broadband broadband penetration increase penetration increase (p. 478) (p. 478)

  • KOUTROUMPIS

KOUTROUMPIS,

, Pantelis Pantelis, , “ “The Economic Impact of The Economic Impact of Broadband on Growth: A Broadband on Growth: A Simultaneous Approach Simultaneous Approach”, ”,

Telecommunications Policy Telecommunications Policy , num. , num. 33, pp.471 33, pp.471– –485, Elsevier, Oct. 485, Elsevier, Oct. 2009 2009

slide-13
SLIDE 13

(Koutroumpis,2009) (Koutroumpis,2009) Reference Model Reference Model

  • Uses

Uses 4 4 simultaneous equations: simultaneous equations:

  • Aggregated Production Function

Aggregated Production Function – – GDP. GDP.

ln ln( (GDP GDPt

t) =

) = β β β β β β β βP0

P0 +

+ β β β β β β β βP1

P1.ln(K

.ln(K t

t )+

)+β β β β β β β βP2

P2.ln(LF

.ln(LF t

t) +

) + β β β β β β β βP3

P3.ln(

.ln(PEN PENt

t) +

) + ε ε ε ε ε ε ε εP

P t t P0 P0 P1 P1 t t P2 P2 t t P3 P3 t t P P

  • Demand for Broadband Infrastructure.

Demand for Broadband Infrastructure.

ln ln( (PEN PENt

t)=

)=β β β β β β β βD

D0 0 +

+β β β β β β β βD

D1 1.ln(GDP

.ln(GDP_

_PER

PER_

_CAPITA

CAPITA t

t ) +

) + β β β β β β β βD

D2 2.(

.(BB BB_

_PRICE

PRICEt

t)

) + + β β β β β β β βD

D3 3.(

.(EDU EDUt

t) +

) +β β β β β β β βD4

D4.(

.(URB URBt

t) +

) + β β β β β β β βD5

D5.(

.(R&D R&Dt

t) +

) + ε ε ε ε ε ε ε εD

D

  • Supply of Broadband Infrastructure.

Supply of Broadband Infrastructure.

ln ln( (BBI BBI ) = ) =β β β β β β β β + +β β β β β β β β .ln(BB .ln(BB PRICE PRICE ) + ) + β β β β β β β β .ln( .ln(REGULATION REGULATION ) + ) +

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 13

ln ln( (BBI BBIt

t) =

) =β β β β β β β βO

O0 0 +

+β β β β β β β βO

O1 1.ln(BB

.ln(BB_

_PRICE

PRICE t

t ) +

) + β β β β β β β βO

O3 3.ln(

.ln(REGULATION REGULATIONt

t) +

) + + + β β β β β β β βO

O2 2.ln(COMPETITION

.ln(COMPETITION_

_TECH

TECH t

t) + +

) + + ε ε ε ε ε ε ε εO

O

  • Broadband Infrastructure Production Function

Broadband Infrastructure Production Function

( )

  • β

β ε

−

  = + ⋅ +    

slide-14
SLIDE 14

(Koutroumpis,2009) (Koutroumpis,2009) Reference Model Reference Model

  • K

K t

t :

: Stock of telecommunications investment.

Stock of telecommunications investment.

  • LF

LF t

t :

: Labor force. Population between 15 and 64 years old.

Labor force. Population between 15 and 64 years old.

  • PEN

PENt

t : Broadband penetration. Number of broadband accesses per

: Broadband penetration. Number of broadband accesses per 100 inhabitants. . 100 inhabitants. .

  • BB_PRICE

BB_PRICE : : Average price of a 1Mbps broadband connection.

Average price of a 1Mbps broadband connection.

  • BB_PRICE

BB_PRICEt

t :

: Average price of a 1Mbps broadband connection.

Average price of a 1Mbps broadband connection.

  • EDU

EDUt

t and

and R&D R&Dt

t :

: Percentage of the GDP spent annually on

Percentage of the GDP spent annually on education and research and development. . education and research and development. .

  • URB

URBt

t :

: Percentage of the population living in areas with

Percentage of the population living in areas with demographic density ≥ 500 demographic density ≥ 500 inhab inhab./km ./km2

2.

.

  • BBI

BBIt

t :

: Broadband annual investment. .

Broadband annual investment. .

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 14
  • BBI

BBIt

t :

: Broadband annual investment. .

Broadband annual investment. .

  • COMPETITION_TECH

COMPETITION_TECHt

t :

: Herfindahl

Herfindahl-

  • Hirschman index, HIRSCHMAN

Hirschman index, HIRSCHMAN (1964), for the competition among broadband technologies as DSL, (1964), for the competition among broadband technologies as DSL, WiFi WiFi, , WiMAX WiMAX, fiber optics, 3G cellular and others , fiber optics, 3G cellular and others

  • REGULATIOn

REGULATIOnt

t :

: Percentage of broadband accesses offered through

Percentage of broadband accesses offered through unbundling unbundling

slide-15
SLIDE 15

Models Used in the Present Study Models Used in the Present Study

  • Used

Used 6 models 6 models ( (3 3 for the for the GDP GDP and and 3 3 for for the the GDP per Capita GDP per Capita) )

  • Model

Model 1 1/ /2 2: : GDP/ GDP/GDP per Capita GDP per Capita: : Without price variable.

Without price variable.

Model Model 1 1/ /2 2: : GDP/ GDP/GDP per Capita GDP per Capita: : Without price variable.

Without price variable.

  • Model

Model 3 3/ /4 4 : : GDP/ GDP/GDP per Capita GDP per Capita: : With price variable. With price variable.

  • Model

Model 5 5/ /6 6 : : GDP/ GDP/GDP per Capita GDP per Capita: : With price variable With price variable but but without without the variables regarding the variables regarding demographic demographic density density and and education education level because lack of expected level because lack of expected results: results:

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 15

results: results:

  • Possible cause: price

Possible cause: price has bigger impact compared to has bigger impact compared to

  • ther variables
  • ther variables (price elasticity in Brazil from

(price elasticity in Brazil from -

  • 1

1 to to

  • 3,36, Ávila (2008)),

3,36, Ávila (2008)), reducing the importance of reducing the importance of education education and and demographic density demographic density in the model. in the model.

slide-16
SLIDE 16

Model Equations Model Equations

  • Aggregated Production Function

Aggregated Production Function – – GDP: GDP:

  • ln

ln( (GDP GDPt

t) =

) = β β β β β β β βP0

P0 +

+ β β β β β β β βP1

P1.ln(

.ln(INVEST_BB INVEST_BBt

t) +

) + β β β β β β β βP2

P2.ln(POP_15_YR_8_YR_EDU

.ln(POP_15_YR_8_YR_EDUt

t) +

) + β β β β β β β βP3

P3.ln(DENS_BB

.ln(DENS_BB t

t) +

) + ε ε ε ε ε ε ε εP

P

  • ln

ln( (GDP_PER_CAPTA GDP_PER_CAPTAt

t) =

) = β β β β β β β βP0

P0 +

+ β β β β β β β βP1

P1.ln(

.ln(INVEST_BB INVEST_BBt

t)

)

  • ln

ln( (GDP_PER_CAPTA GDP_PER_CAPTAt

t) =

) = β β β β β β β βP0

P0 +

+ β β β β β β β βP1

P1.ln(

.ln(INVEST_BB INVEST_BBt

t)

) + +β β β β β β β βP2

P2.ln(POP_15_YR_8_YR_EDU

.ln(POP_15_YR_8_YR_EDU t

t) +

) + β β β β β β β βP3

P3.ln(DENS_BB

.ln(DENS_BB t

t) +

) + ε ε ε ε ε ε ε εP

P

  • Demand for Broadband Infrastructure:

Demand for Broadband Infrastructure:

ln ln(DENS (DENS_

_BB

BB t

t) =

) = β β β β β β β βD0

D0 +

+β β β β β β β βD1

D1.ln(

.ln(GDP GDP_

_PER

PER_

_CAPTA

CAPTAt

t) +

) + β β β β β β β βD4

D4.ln(

.ln(PRICE PRICEt

t)

) + + β β β β β β β βD2

D2.(

.(PERCENT_EDU PERCENT_EDUt

t) +

) + β β β β β β β βD3

D3.ln(POP_50K

.ln(POP_50K t

t) +

) + ε ε ε ε ε ε ε εD

D

  • Supply of Broadband Infrastructure:

Supply of Broadband Infrastructure:

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 16
  • Supply of Broadband Infrastructure:

Supply of Broadband Infrastructure:

ln ln( (INVEST_BB INVEST_BBt

t) =

) = β β β β β β β βO0

O0 +

+ β β β β β β β βO1

O1.ln(

.ln(REV_BB REV_BBt

t) +

) + β β β β β β β βO2

O2.ln(

.ln(PRICE PRICEt

t)+

)+ ε ε ε ε ε ε ε εO

O

  • Broadband Infrastructure Production Function:

Broadband Infrastructure Production Function:

  • ε

β β + ⋅ + =        

−
slide-17
SLIDE 17

Variables Variables

  • GDP

GDPt

t ,

, GDP_PER_CAPTA GDP_PER_CAPTAt

t:

: GDP and GDP per GDP and GDP per capta capta per State from per State from 2000 to 2008 2000 to 2008. . Source IBGE Source IBGE. . Obs. :

  • Obs. :

2008 data estimated. 2008 data estimated.

  • INVEST_BB

INVEST_BB and and REV_BB REV_BB : : Broadband annual Broadband annual

  • INVEST_BB

INVEST_BBt

t and

and REV_BB REV_BBt

t :

: Broadband annual Broadband annual investments and Gross annual operational revenue investments and Gross annual operational revenue (2000 to 2008) (2000 to 2008) of the broadband providers. Data

  • f the broadband providers. Data

aggregated for the whole country. aggregated for the whole country. Source: ANATEL. Source: ANATEL.

  • POP_15_YR_8_YR_EDUt

POP_15_YR_8_YR_EDUt , , PERCENT_EDU PERCENT_EDU : :

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 17
  • POP_15_YR_8_YR_EDUt

POP_15_YR_8_YR_EDUt t

t ,

, PERCENT_EDU PERCENT_EDUt

t :

: population and percentage of the State population population and percentage of the State population at least 15 years old and with 8 years or more of at least 15 years old and with 8 years or more of complete education. complete education. Source: IBGE PNAD Source: IBGE PNAD. .

slide-18
SLIDE 18

Variables (cont.) Variables (cont.)

  • POP_50K

POP_50K t

t:

: Share of the State population living in Share of the State population living in cities with at least 50.000 inhabitants. cities with at least 50.000 inhabitants. Proxy for

Proxy for population concentration, replacing the variable used by population concentration, replacing the variable used by Koutroumpis Koutroumpis (2009) of the share of the population living (2009) of the share of the population living Koutroumpis Koutroumpis (2009) of the share of the population living (2009) of the share of the population living in areas where density ≥ 500 in areas where density ≥ 500 inhab inhab./ km ./ km2

2.

. Source: IBGE. Source: IBGE.

  • PRICE

PRICEt

t:

: Estimated Estimated average price charged for the average price charged for the broadband access. broadband access. Source: Estimated from Source: Estimated from CETIC.BR CETIC.BR anual anual surveys from 2005 to 2008. surveys from 2005 to 2008.

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 18
  • DENS_BB

DENS_BBt

t :

: Broadband density. Number of Broadband density. Number of accesses per 1000 inhabitants, for each State, accesses per 1000 inhabitants, for each State, between 2000 and 2008. between 2000 and 2008. Source Source Anatel Anatel (SICI). (SICI). Obs.: data from 2000 to 2006 estimated. Obs.: data from 2000 to 2006 estimated.

slide-19
SLIDE 19

+$

  • *!

;<=

;<= &! 3= 8>

Variables: Variables: GDP per Capita X Broadband Penetration GDP per Capita X Broadband Penetration

(?) +?)% , (?$)#

$ + +$ & +

8>

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 19

$

  • ,.

,.+? ,.? ,."? ,.#? ,.$?

  • !

<=&67 ! ,.!

slide-20
SLIDE 20

+% +

  • *!

;1

;1 &! 3= 8> !

Variables: Variables: Broadband Penetration X Education Broadband Penetration X Education

(?#+"?)+ , (?$$%

%

  • +

+ +# +% & +

8> ! @ !("?A+"?)+* , (?$$%

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 20
  • #

*+ * *" *# *$ *%

  • !

&1&>+$2B> 2=1 !

slide-21
SLIDE 21

$ % ! *!

!;1

!;1 &! 3= 8> !

Variables: Variables: ln ln(Broadband Penetration) X Education (Broadband Penetration) X Education

(+"*)+ "*% ,C(*$$$

  • +
  • "

# &! +

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 21

"

  • + *+

* *" *# *$ *% !

/>&1&>+$2 B>2=1 !

slide-22
SLIDE 22

,."$? ,.#? ,.#$? ! ,.!

<=;1

<=;1 &! 3= 8> !

Variables: Variables: GDP per Capita X Education GDP per Capita X Education

($+?%?+) , (?$)

,.+? ,.+$? ,.? ,.$? ,."? ,."$? & ! ,

8> ! !(%?#A%?+)* , (?$)

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 22

,. ,.$? ,.+?

  • *+

* *" *# *$ *% * <= &1&>+$2B >2=1 !

slide-23
SLIDE 23

Dependent Dependent Variables Variables Coef. Coef. Model Model 1 1 Model Model 3 3 Model Model 5 5 GMM GMM 3SLS 3SLS GMM GMM 3SLS 3SLS GMM GMM 3SLS 3SLS

Aggregated Production Function Aggregated Production Function – – GDP Per Capita GDP Per Capita INVEST_BB INVEST_BBt

t

β β β β β β β βP1

P1

0,384 0,384(48,365)

(48,365) 0,381

0,381(26,139)

(26,139)

0,386 0,386(53,844)

(53,844)

0,392 0,392(29,163)

(29,163) 0,381

0,381(51,346)

(51,346)

0,397 0,397(29,672)

(29,672)

POP POP_

_15

15_

_YR_8

YR_8_

_YR

YR_

_EDU

EDUt

t β

β β β β β β βP2

P2 1,010

1,010(86,521)

(86,521) 1,095

1,095(47,894)

(47,894)

1,103 1,103(96,447)

(96,447)

1,093 1,093(51,880)

(51,880) 1,116

1,116(91,913)

(91,913)

1,084 1,084(51,745)

(51,745)

DENS_BB DENS_BB t

t

β β β β β β β βP3

P3

0,133 0,133(7,806)

(7,806)

0,180 0,180(8,670)

(8,670)

0,097 0,097(6,708)

(6,708)

0,104 0,104(5,613)

(5,613)

0,038 0,038(2,228)

(2,228) * *

0,109 0,109(5,860)

(5,860)

Demand for Broadband Demand for Broadband

1,088 1,088 0,784 0,784 1,713 1,713 1,585 1,585 1,453 1,453 1,244 1,244

Regression Results: GDP Regression Results: GDP

Demand for Broadband Demand for Broadband GDP_PER_CAPITA GDP_PER_CAPITAt

t

β β β β β β β βD1

D1

1,088 1,088(6,998)

(6,998)

0,784 0,784(5,796)

(5,796)

1,713 1,713(20,328)

(20,328)

1,585 1,585(18,552)

(18,552) 1,453

1,453(42,766)

(42,766)

1,244 1,244(46,685)

(46,685)

PERCENT_EDU PERCENT_EDUt

t

β β β β β β β βD2

D2

0,149 0,149(17,000)

(17,000)

0,149 0,149(16,465)

(16,465)
  • 0,029

0,029(

(-
  • 4,368)
4,368)
  • 0,022

0,022(

(-
  • 2,471)
2,471) – – – –

POP_50K POP_50K t

t

β β β β β β β βD3

D3 -
  • 3,298

3,298(

(-
  • 11,347)
11,347) -
  • 2,622

2,622(

(-
  • 9,661)
9,661) -
  • 0,374

0,374*

*( (-
  • 2,325)
2,325) -
  • 0,458

0,458(

(-
  • 2,512)
2,512) – – – –

PRICE PRICE t

t

β β β β β β β βD4

D4 – – – –
  • 2,111

2,111(

(-
  • 29,581)
29,581) -
  • 1,871

1,871(

(-
  • 22,634)
22,634) -
  • 2,128

2,128(

(-
  • 38,059)
38,059) -
  • 1,789

1,789(

(-
  • 39,769)
39,769)

Broadband Supply Broadband Supply Constant Constant β β β β β β β βO0

O0

4,710 4,710(9,126)

(9,126)

3,377 3,377(6,925)

(6,925) – – – – – – – –

REV_BB REV_BBt

t

β β β β β β β βO1

O1 0,739

0,739(33,039)

(33,039) 0,797

0,797(37,704)

(37,704) 0,890

0,890(309,895)

(309,895)

0,901 0,901(211,142)

(211,142) 0,889

0,889(278,367)

(278,367)

0,902 0,902(211,302)

(211,302)

PRICE PRICE t

t

β β β β β β β βO3

O3 – – – –

0,246 0,246(19,626)

(19,626)

0,196 0,196(10,183)

(10,183)

0,248 0,248(17,776)

(17,776)

0,191 0,191(9,926)

(9,926)

Broadband Infrastructure Production Function Broadband Infrastructure Production Function

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 23

Broadband Infrastructure Production Function Broadband Infrastructure Production Function INVEST_BB INVEST_BBt

t

β β β β β β β βPBL1

PBL1

0,023 0,023(19,384)

(19,384) 0,022

0,022(16,399)

(16,399)

0,022 0,022(21,954)

(21,954)

0,022 0,022(16,121)

(16,121)

0,022 0,022(21,124)

(21,124)

0,021 0,021(16,090)

(16,090)

R R2

2

Aggregated Production Aggregated Production

0,944 0,944 0,940 0,940 0,944 0,944 0,945 0,945 0,940 0,940 0,945 0,945

Demand Demand

0,641 0,641 0,652 0,652 0,854 0,854 0,873 0,873 0,813 0,813 0,865 0,865

Supply Supply

0,853 0,853 0,856 0,856 0,838 0,838 0,853 0,853 0,839 0,839 0,854 0,854

Infrastructure Production Infrastructure Production

  • 0,031
0,031
  • 0,025
0,025
  • 0,025
0,025
  • 0,025
0,025
  • 0,025
0,025
  • 0.026
0.026 !!5+!60 *785+!0 !!5+!60 *785+!0
slide-24
SLIDE 24

Dependent Dependent Variables Variables Coef. Coef. Model 2 Model 2 Model Model 4 4 Model Model 6 6 GMM GMM 3SLS 3SLS GMM GMM 3SLS 3SLS GMM GMM 3SLS 3SLS

Aggregated Production Function Aggregated Production Function – – GDP Per Capita GDP Per Capita INVEST_BB INVEST_BBt

t

β β β β β β β βP1

P1 0,301

0,301(23,248)

(23,248) 0,321

0,321(16,891)

(16,891)

0,291 0,291(23,375)

(23,375)

0,314 0,314(16,462)

(16,462)

0,307 0,307(27,641)

(27,641)

0,327 0,327(18,751)

(18,751)

POP POP_

_15

15_

_YR_8

YR_8_

_YR

YR_

_EDU

EDUt

t β

β β β β β β βP2

P2 0,126

0,126(6,226)

(6,226)

0,088 0,088(2,989)

(2,989)

0,146 0,146(7,505)

(7,505)

0,118 0,118(3,952)

(3,952)

0,136 0,136(7,663)

(7,663)

0,104 0,104(3,809)

(3,809)

DENS_BB DENS_BB t

t

β β β β β β β βP3

P3 0,321

0,321(13,731)

(13,731) 0,362

0,362(12,664)

(12,664)

0,260 0,260(12,980)

(12,980)

0,238 0,238(9,432)

(9,432)

0,196 0,196(8,353)

(8,353)

0,203 0,203(8,015)

(8,015)

Demand for Broadband Demand for Broadband

Regression Results: GDP per Capita Regression Results: GDP per Capita

Demand for Broadband Demand for Broadband GDP_PER_CAPITA GDP_PER_CAPITAt

t

β β β β β β β βD1

D1 1,003

1,003(6,028)

(6,028)

0,488 0,488(3,235)

(3,235)

2,109 2,109(17,375)

(17,375)

2,005 2,005(17,337)

(17,337)

1,319 1,319(39,890)

(39,890)

1,340 1,340(44,020)

(44,020)

PERCENT_EDU PERCENT_EDUt

t

β β β β β β β βD2

D2 0,154

0,154(17,625)

(17,625) 0,164

0,164(17,734)

(17,734) -
  • 0,056

0,056(

(-
  • 5,416)
5,416)
  • 0,045

0,045(

(-
  • 4,155)
4,155) – – – –

POP_50K POP_50K t

t

β β β β β β β βD3

D3 -
  • 3,158

3,158(

(-
  • 9,920)
9,920) -
  • 2,090

2,090(

(-
  • 6,915)
6,915) -
  • 0,972

0,972(

(-
  • 4,275)
4,275)
  • 0,924

0,924(

(-
  • 4,299)
4,299) – – – –

PRICE PRICE t

t

β β β β β β β βD4

D4 – – – –
  • 2,156

2,156(

(-
  • 22,467)
22,467) -
  • 2,071

2,071(

(-
  • 21,273)
21,273) -
  • 1,928

1,928 (

(-
  • 35,367)
35,367) -
  • 1,950

1,950(

(-
  • 37,986)
37,986)

Broadband Supply Broadband Supply Constant Constant β β β β β β β βO0

O0 4,636

4,636(8,835)

(8,835)

2,961 2,961(5,985)

(5,985) – – – – – – – –

REV_BB REV_BBt

t

β β β β β β β βO1

O1 0,742

0,742(32,631)

(32,631) 0,815

0,815(38,014)

(38,014) 0,896

0,896(275,698)

(275,698) 0,901

0,901(209,549)

(209,549)

0,89 0,89(245,745)

(245,745)

0,900 0,900(208,769)

(208,769)

PRICE PRICE t

t

β β β β β β β βO3

O3 – – – –

0,216 0,216(15,272)

(15,272)

0,192 0,192(9,909)

(9,909)

0,207 0,207(12,835)

(12,835)

0,199 0,199 (10,221)

(10,221) Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 24

PRICE PRICE t

t

β β β β β β β βO3

O3 – – – –

0,216 0,216(15,272)

(15,272)

0,192 0,192(9,909)

(9,909)

0,207 0,207(12,835)

(12,835)

0,199 0,199 (10,221)

(10,221)

Broadband Infrastructure Production Function Broadband Infrastructure Production Function INVEST_BB INVEST_BBt

t

β β β β β β β βPBL1

PBL1 0,024

0,024(19,379)

(19,379) 0,022

0,022(16,391)

(16,391)

0,021 0,021(20,657)

(20,657)

0,022 0,022(16,424)

(16,424)

0,023 0,023(22,062)

(22,062)

0,022 0,022(16,362)

(16,362)

R R2

2

Aggregated Production Aggregated Production 0,420 0,420 0,389 0,389 0,447 0,447 0,465 0,465 0,456 0,456 0,463 0,463 Demand Demand 0,642 0,642 0,642 0,642 0,846 0,846 0,856 0,856 0,852 0,852 0,850 0,850 Supply Supply 0,853 0,853 0,854 0,854 0,848 0,848 0,854 0,854 0,849 0,849 0,853 0,853 Infrastructure Production Infrastructure Production

  • 0,033

0,033

  • 0,025

0,025

  • 0,028

0,028

  • 0,025

0,025

  • 0,027

0,027

  • 0,025

0,025

All coefficients at 1% significance level. All coefficients at 1% significance level.

slide-25
SLIDE 25

Results: Comments Results: Comments

  • High economic impact. Possible reasons:

High economic impact. Possible reasons:

  • Data problems: estimated values instead of

Data problems: estimated values instead of real ones: real ones: Ex.: Broadband Penetration per Ex.: Broadband Penetration per State; Broadband prices. State; Broadband prices. State; Broadband prices. State; Broadband prices.

  • Historical series too recent (few samples):

Historical series too recent (few samples):

  • Annual data from 2000 to 2008 only.

Annual data from 2000 to 2008 only.

  • Low broadband penetration in Brazil:

Low broadband penetration in Brazil:

  • 2007: 45.8 accesses/100

2007: 45.8 accesses/100 inhab inhab. . High growth of

High growth of

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 25
  • 2007: 45.8 accesses/100

2007: 45.8 accesses/100 inhab inhab. .

  • 2008: 59.1 accesses/100

2008: 59.1 accesses/100 inhab inhab. .

  • Important omitted variables:

Important omitted variables:

  • Ex.: Backhaul availability data

Ex.: Backhaul availability data

High growth of High growth of penetration penetration level ( level ( ≈ ≈ ≈ ≈ ≈ ≈ ≈ ≈ 30% ) 30% )

  • ver a small
  • ver a small

base of base of subscribers. subscribers.

slide-26
SLIDE 26

Thank You! Thank You! Thank You! Thank You!

Hildebrando Hildebrando R.

  • R. Macedo

Macedo -

  • ANATEL

ANATEL hmacedo@anatel.gov.br hmacedo@anatel.gov.br

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 26

Alexandre Alexandre X.Ywata X.Ywata de de Carvalho Carvalho -

  • IPEA

IPEA alexandre.ywata@ipea.gov.br alexandre.ywata@ipea.gov.br

slide-27
SLIDE 27

Additional Additional Additional Additional Slides Slides

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 27

Slides Slides

slide-28
SLIDE 28

Introduction Introduction

  • Broadband networks

Broadband networks became important became important development development aid tools for the countries, allowing aid tools for the countries, allowing to transform the existing economic activities as to transform the existing economic activities as well to create new ones. well to create new ones. well to create new ones. well to create new ones.

  • Several countries, like

Several countries, like USA, Brazil, Australia, Spain USA, Brazil, Australia, Spain and others started to implement and others started to implement public policies public policies aiming increase the broadband penetration. aiming increase the broadband penetration.

  • Those initiatives had fundament on

Those initiatives had fundament on economic economic

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 28
  • Those initiatives had fundament on

Those initiatives had fundament on economic economic impact studies impact studies showing the importance of showing the importance of broadband networks broadband networks to boost economic to boost economic development. development.

slide-29
SLIDE 29

Introduction and Motivation Introduction and Motivation

  • Several countries, like

Several countries, like USA, Brazil, Australia, Spain USA, Brazil, Australia, Spain and others started to implement and others started to implement public policies public policies aiming increase the broadband penetration. aiming increase the broadband penetration.

  • Those had fundament on

Those had fundament on economic impact studies economic impact studies

  • Those had fundament on

Those had fundament on economic impact studies economic impact studies showing the importance of showing the importance of broadband networks broadband networks to to boost economic boost economic development. development.

  • The

The proper proper design of national design of national public policies public policies to to expand broadband penetration need expand broadband penetration need individual individual

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 29

expand broadband penetration need expand broadband penetration need individual individual Economic Impact Economic Impact studies for each country to take studies for each country to take into account its particularities. into account its particularities.

  • This was the motivation to study the broadband

This was the motivation to study the broadband economic impact in Brazil. economic impact in Brazil.

slide-30
SLIDE 30
  • This study tried to address these

This study tried to address these questions, for the questions, for the Brazilian Brazilian case: case:

  • Identify if there is a positive link between

Identify if there is a positive link between

Objectives of the Study Objectives of the Study

  • Identify if there is a positive link between

Identify if there is a positive link between the increase in the broadband penetration the increase in the broadband penetration and the local economic development. and the local economic development.

  • Evaluate the broadband economic impact

Evaluate the broadband economic impact

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 30
  • Evaluate the broadband economic impact

Evaluate the broadband economic impact in Brazil using official statistics data from in Brazil using official statistics data from Anatel Anatel (the Brazilian telecom regulatory (the Brazilian telecom regulatory agency) agency)

slide-31
SLIDE 31

References References

  • World Bank study

World Bank study ( (Qiang Qiang and others, 2009, and others, 2009,

  • p. 35 to 50) *
  • p. 35 to 50) *

Increase of Increase of 1,38 1,38 p.p. on the p.p. on the

  • Increase of

Increase of 1,38 1,38 p.p. on the p.p. on the GDP per capita growth rate GDP per capita growth rate for each for each 10 10 p.p. increase in p.p. increase in the broadband penetration the broadband penetration for developing countries for developing countries

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 31

for developing countries for developing countries (p. 45) . (p. 45) .

  • *

* QIANG QIANG, Christine , Christine Zhen Zhen-

  • Wei

Wei; ; ROSSOTTO ROSSOTTO, Carlo M. e , Carlo M. e KIMURA, KIMURA, Kaoru Kaoru, “ , “Economic Economic Impacts Impacts of

  • f Broadband

Broadband”, em ”, em ICAD2009 ICAD2009 -

  • Information

Information and and Communications for Communications for Development Development, Capítulo 3 : , Capítulo 3 : Extending Extending Reach Reach and and Increasing Increasing Impact Impact, pp. 35 a 50, , pp. 35 a 50, The The World World Bank Bank, Washington,DC, , Washington,DC, 2009 2009

slide-32
SLIDE 32

References References

  • Crandall and others (2007)

Crandall and others (2007)

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 32

Crandall and others (2007) Crandall and others (2007)

  • Increase between

Increase between 0,2 to 0,3 0,2 to 0,3 p.p. in the p.p. in the employment level for each employment level for each 1 1 p.p. increase in p.p. increase in the broadband penetration the broadband penetration (p. 2) (p. 2)

slide-33
SLIDE 33

References References

  • Koutsky

Koutsky and Ford and Ford (2005) (2005)

  • KOUTSKY

KOUTSKY , , Thomas Thomas M

  • M. e

. e FORD , FORD ,George George S S., " ., "Broadband and Broadband and Economic Development: A Municipal Case Study from Florida Economic Development: A Municipal Case Study from Florida" , " , Review of Urban & Regional Development Studies Review of Urban & Regional Development Studies, Vol. 17, No. 3, , Vol. 17, No. 3,

  • pp. 219
  • pp. 219-
  • 229, 2005,

229, 2005, Journal of the Applied Regional Science Journal of the Applied Regional Science

  • pp. 219
  • pp. 219-
  • 229, 2005,

229, 2005, Journal of the Applied Regional Science Journal of the Applied Regional Science C Conference

  • nference
  • Approximately

Approximately 100%

100% increase in the

increase in the economic economic activity activity of

  • f Lake County, Florida (USA)

Lake County, Florida (USA) after after public investment in the deployment of public investment in the deployment of extensive fiber optics network. extensive fiber optics network.

  • This when compared to similar localities that

This when compared to similar localities that

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 33
  • This when compared to similar localities that

This when compared to similar localities that had not had not such investment. such investment.

slide-34
SLIDE 34
  • Evaluated broadband economic impact

Evaluated broadband economic impact

  • n
  • n 22 OECD

22 OECD, , countries with data from countries with data from 2002 to 2007 2002 to 2007. .

(Koutroumpis,2009) (Koutroumpis,2009) Reference Study Reference Study

2002 to 2007 2002 to 2007. .

  • Reproduced

Reproduced Röller Röller e e Waverman Waverman (2001 (2001) ) study which evaluated economic impact study which evaluated economic impact

  • f fixed line telephone networks.
  • f fixed line telephone networks.
  • Results :

Results :

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 34
  • Results :

Results : Each Each 1p.p. 1p.p. increase in the broadband increase in the broadband penetration was related to about penetration was related to about 0,025p.p. 0,025p.p. GDP increase (p. 477). GDP increase (p. 477).

slide-35
SLIDE 35

Reference Model Reference Model

  • Replicated

Replicated (Koutroumpis,2009) (Koutroumpis,2009) model model with data from Brazil with data from Brazil

  • Model:

Model: Simultaneous equations of supply Simultaneous equations of supply and demand with endogenous variables. and demand with endogenous variables.

  • Model:

Model: Simultaneous equations of supply Simultaneous equations of supply and demand with endogenous variables. and demand with endogenous variables.

  • Justification:

Justification: To capture the simultaneous To capture the simultaneous effect of telecom investment on the effect of telecom investment on the economy: economy:

  • Investment in telecommunications

Investment in telecommunications lead to lead to increase in the economic development increase in the economic development

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 35

increase in the economic development increase in the economic development At the same time: At the same time:

  • Economic development

Economic development of the country

  • f the country

increases demand for increases demand for investment in investment in telecommunications infrastructure. telecommunications infrastructure.

slide-36
SLIDE 36

%* * *

'!

  • D&0&

Broadband Accesses per Speed Broadband Accesses per Speed

  • +
  • "
# $ %
  • )
+

* +* * "* #* $*

  • D& '
Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 36
  • +
  • "
# $ %
  • )
+ %#E $* ""*" "*% "#*% %*) *+ +)*$ +*% ++*% ++* +*# %#E$+E +$*# %*$ %%*+ %+* #$* $*% $+* $)*) "*+ *% +*+ $+E- $* "*+ *% "* * %*$ *% $*$ #*" ##*) #)*%
  • "#-
*+ * *$ *# *) *% +* +*% *" +#* +*% :"#- *) *# *" * * * * *" * * *" 8F80> '! + + + + + + + + + + +

*

2

slide-37
SLIDE 37

Broadband Accesses per Technology Broadband Accesses per Technology

$ %

  • '!
  • D&8
  • +
  • "

#

  • D& '!
Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 37
  • +
  • "
# $ %
  • )
+ F8 #*+ +)*)+ +*+ *%) "*"% )*%$ +%*# +*#" +*$$ ++* )* &> +*$ %*$% %+*)% $#*)$ %*)# %+*% %%*" %%*+ %$*+" %%*$ $)*" B ++* *+% $*" "*$) *) * +*$ #*) *) " $*+" =- *$ +* $*)" *"# "*$ "*$# +#*+$ +$*) +*%) +*#" "*+% & *)$ +*%% )* %*#" *# * +*" +*#) +* +* *)"
  • 2
slide-38
SLIDE 38

Estimated Data Used Estimated Data Used

  • Some Important Data had to be

Some Important Data had to be Estimated: Estimated: Estimated: Estimated:

  • Broadband penetration

Broadband penetration disaggregated by State from 2000 disaggregated by State from 2000 to 2006. to 2006.

  • GDP per State in 2008

GDP per State in 2008

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 38
  • GDP per State in 2008

GDP per State in 2008

  • Broadband Prices

Broadband Prices

slide-39
SLIDE 39

Data Estimation: Broadband Penetration Data Estimation: Broadband Penetration per State from 2000 to 2006 per State from 2000 to 2006

S Share hare of each

  • f each
  • ne of the
  • ne of the 27

27

A Appr pproximation:

  • ximation:

Share Share of each

  • f each

S Share hare of each

  • f each
  • ne of the
  • ne of the 27

27 Brazilian States Brazilian States

  • n the total
  • n the total

internet accesses internet accesses

≈ ≈ ≈ ≈ ≈ ≈ ≈ ≈

Share Share of each

  • f each
  • ne of the
  • ne of the 27

27 Brazilian States Brazilian States

  • n the total
  • n the total

broadband broadband

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 39

internet accesses internet accesses

(broadband or (broadband or not) not) in the

in the country country

broadband broadband accesses accesses in the in the country. country.

slide-40
SLIDE 40

2007 2007 2008 2008

STATES STATES´ ´ SHARES ON THE COUNTRY SHARES ON THE COUNTRY´ ´S S TOTAL TOTAL NUMBER OF: NUMBER OF:

STATE STATE

BROADBAND BROADBAND ACCESSES ACCESSES

(4th Quarter (4th Quarter 2007) 2007) (%) (ANATEL (%) (ANATEL-
  • SICI)
SICI)

INTERNET INTERNET ACCESSES ACCESSES

(%) (IBGE (%) (IBGE -
  • PNAD 2007)
PNAD 2007)

BROADBAND BROADBAND ACCESSES ACCESSES

(4th Quarter (4th Quarter 2008) 2008) (%) (ANATEL (%) (ANATEL-
  • SICI)
SICI)

INTERNET INTERNET ACCESSES ACCESSES

(%) (IBGE (%) (IBGE -
  • PNAD 2008)
PNAD 2008)

AC AC 0.333 0.333 0.184 0.184 0.155 0.155 0.224 0.224 AL AL 0.371 0.371 0.473 0.473 0.357 0.357 0.605 0.605 AM AM 0.477 0.477 0.608 0.608 0.491 0.491 0.701 0.701 AP AP 0.036 0.036 0.119 0.119 0.048 0.048 0.107 0.107 * * BA BA 2.774 2.774 3.576 3.576 2.876 2.876 4.162 4.162 * * BA BA 2.774 2.774 3.576 3.576 2.876 2.876 4.162 4.162 * * CE CE 1.789 1.789 1.559 1.559 1.641 1.641 1.896 1.896 * * DF DF 3.860 3.860 2.511 2.511 4.348 4.348 2.471 2.471 ES ES 1.345 1.345 2.074 2.074 1.577 1.577 2.040 2.040 * * GO GO 2.467 2.467 2.249 2.249 2.556 2.556 2.366 2.366 MA MA 0.539 0.539 0.713 0.713 0.519 0.519 0.916 0.916 * * MG MG 8.320 8.320 9.774 9.774 9.040 9.040 10.372 10.372 MS MS 0.831 0.831 1.027 1.027 1.213 1.213 1.060 1.060 MT MT 0.408 0.408 0.959 0.959 1.313 1.313 1.385 1.385 PA PA 0.975 0.975 1.033 1.033 0.984 0.984 1.161 1.161 PB PB 0.494 0.494 0.743 0.743 0.620 0.620 0.960 0.960 * * PE PE 1.731 1.731 2.049 2.049 1.715 1.715 2.160 2.160 PI PI 0.393 0.393 0.468 0.468 0.397 0.397 0.504 0.504

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 40

PI PI 0.393 0.393 0.468 0.468 0.397 0.397 0.504 0.504 * * PR PR 6.984 6.984 7.214 7.214 7.816 7.816 7.366 7.366 * * RJ RJ 11.151 11.151 13.160 13.160 11.087 11.087 12.774 12.774 RN RN 0.498 0.498 0.819 0.819 0.536 0.536 0.870 0.870 RO RO 0.941 0.941 0.488 0.488 0.435 0.435 0.489 0.489 RR RR 0.023 0.023 0.115 0.115 0.023 0.023 0.118 0.118 * * RS RS 7.394 7.394 6.808 6.808 6.884 6.884 6.594 6.594 * * SC SC 5.514 5.514 4.737 4.737 4.434 4.434 4.774 4.774 SE SE 0.452 0.452 0.586 0.586 0.441 0.441 0.643 0.643 * * SP SP 39.535 39.535 35.684 35.684 38.151 38.151 32.975 32.975 TO TO 0.367 0.367 0.270 0.270 0.343 0.343 0.311 0.311

slide-41
SLIDE 41

2007 2007 2008 2008

A A B B C C D D

STATE STATE STATES STATES´ ´ SHARES ON SHARES ON THE COUNTRY THE COUNTRY´ ´S S TOTAL TOTAL BROADBAND BROADBAND ACCESSES (%) ACCESSES (%) STATES STATES´ ´ SHARES ON SHARES ON THE COUNTRY THE COUNTRY´ ´S S TOTAL TOTAL INTERNET INTERNET ACCESSES (%) ACCESSES (%) STATES STATES´ ´ SHARES ON SHARES ON THE COUNTRY THE COUNTRY´ ´S S TOTAL TOTAL BROADBAND BROADBAND ACCESSES (%) ACCESSES (%) STATES STATES´ ´ SHARES ON SHARES ON THE COUNTRY THE COUNTRY´ ´S S TOTAL TOTAL INTERNET INTERNET ACCESSES (%) ACCESSES (%)

BA BA

2.774 2.774 3.576 3.576 2.876 2.876 4.162 4.162

* * BA

BA

2.774 2.774 3.576 3.576 2.876 2.876 4.162 4.162

* * CE

CE

1.789 1.789 1.559 1.559 1.641 1.641 1.896 1.896

* * DF

DF

3.860 3.860 2.511 2.511 4.348 4.348 2.471 2.471

* * GO

GO

2.467 2.467 2.249 2.249 2.556 2.556 2.366 2.366

* * MG

MG

8.320 8.320 9.774 9.774 9.040 9.040 10.372 10.372

* * PE

PE

1.731 1.731 2.049 2.049 1.715 1.715 2.160 2.160

* * PR

PR

6.984 6.984 7.214 7.214 7.816 7.816 7.366 7.366

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 41

* * PR

PR

6.984 6.984 7.214 7.214 7.816 7.816 7.366 7.366

* * RJ

RJ

11.151 11.151 13.160 13.160 11.087 11.087 12.774 12.774

* * RS

RS

7.394 7.394 6.808 6.808 6.884 6.884 6.594 6.594

* * SC

SC

5.514 5.514 4.737 4.737 4.434 4.434 4.774 4.774

* * SP

SP

39.535 39.535 35.684 35.684 38.151 38.151 32.975 32.975

slide-42
SLIDE 42

Estimation Results Estimation Results

  • States

States´ ´ Shares of Shares of Broadband Broadband Accesses Accesses X X X X

  • States

States´ ´ Shares of Shares of Internet Internet Accesses Accesses

  • 2007: R

2007: R2

2 = 0,86

= 0,86

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 42
  • 2008: R

2008: R2

2 = 0,79

= 0,79

slide-43
SLIDE 43

%

  • "

" "# "% " # '! && && =@&&G; &&

  • #

%

  • +

+ +# +% +

  • #

% '!

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 43

Comparison Comparison of

  • f each

each state state share share on

  • n the

the country country5s s total total number number of

  • f broadband

broadband accesses accesses and and on

  • n the

the total total number number of

  • f internet

internet accesses, accesses, for for 2008 2008. . Source Source: : elaborated elaborated by by the the authors authors from from Anatel Anatel and and IBGE IBGE data data. .

  • & ,H-<

, ,& &= 7 0<F 1 =1 1& -8 -&0 ,/ -0 0- &1 ,F 90> 8F 0= 0 ,,

&

slide-44
SLIDE 44

# %

  • "

" "# "% " # '! && && =@&&; &&

  • #

%

  • +

+ +# +% +

  • #

' '!

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 44

Comparison Comparison of

  • f each

each state state share share on

  • n the

the country country5s s total total number number of

  • f broadband

broadband accesses accesses and and on

  • n the

the total total number number of

  • f internet

internet accesses, accesses, for for 2007 2007. . Source Source: : elaborated elaborated by by the the authors authors from from Anatel Anatel and and IBGE IBGE data data. .

  • & ,H -< , ,& &=7 0<F 1 =1 1&-8 -& 0 ,/ -0

0- &1,F 9 0> 8F0= 0 ,, &

slide-45
SLIDE 45

2008 GDP States 2008 GDP States´ ´ Estimates Estimates

  • Only

Only national national 2008 GDP available but not 2008 GDP available but not GDP per State. GDP per State.

  • To estimate the GDP per

To estimate the GDP per State State of

  • f 2008

2008: :

  • To estimate the GDP per

To estimate the GDP per State State of

  • f 2008

2008: :

  • It was assumed that each

It was assumed that each State State would retain would retain the same national GDP share of the same national GDP share of 2007 2007. .

  • Hypothesis reinforced when compared data

Hypothesis reinforced when compared data

  • f
  • f 2006

2006 and and 2007 2007 for the participation of for the participation of

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 45
  • f
  • f 2006

2006 and and 2007 2007 for the participation of for the participation of each State in the national GDP ( each State in the national GDP (similar to similar to the first decimal case the first decimal case) )

slide-46
SLIDE 46

Shares of Each State on the National GDP (%) Shares of Each State on the National GDP (%) (Source: IBGE)

(Source: IBGE)

2002 2002 2003 2003 2004 2004 2005 2005 2006 2006 2007 2007

NORTE NORTE 4.69 4.69 4.78 4.78 4.95 4.95 4.96 4.96 5.06 5.06 5.02 5.02 Rondônia Rondônia 0.53 0.53 0.57 0.57 0.58 0.58 0.60 0.60 0.55 0.55 0.56 0.56 Acre Acre 0.19 0.19 0.19 0.19 0.20 0.20 0.21 0.21 0.20 0.20 0.22 0.22 Amazonas Amazonas 1.47 1.47 1.47 1.47 1.56 1.56 1.55 1.55 1.65 1.65 1.58 1.58 Roraima Roraima 0.16 0.16 0.16 0.16 0.14 0.14 0.15 0.15 0.15 0.15 0.16 0.16 Pará Pará 1.74 1.74 1.75 1.75 1.83 1.83 1.82 1.82 1.87 1.87 1.86 1.86 Amapá Amapá 0.22 0.22 0.20 0.20 0.20 0.20 0.20 0.20 0.22 0.22 0.23 0.23 Tocantins Tocantins 0.38 0.38 0.43 0.43 0.43 0.43 0.42 0.42 0.41 0.41 0.42 0.42 Tocantins Tocantins 0.38 0.38 0.43 0.43 0.43 0.43 0.42 0.42 0.41 0.41 0.42 0.42 NORDESTE NORDESTE 12.96 12.96 12.77 12.77 12.72 12.72 13.07 13.07 13.13 13.13 13.07 13.07 Maranhão Maranhão 1.05 1.05 1.09 1.09 1.11 1.11 1.18 1.18 1.21 1.21 1.19 1.19 Piauí Piauí 0.50 0.50 0.52 0.52 0.51 0.51 0.52 0.52 0.54 0.54 0.53 0.53 Ceará Ceará 1.96 1.96 1.92 1.92 1.90 1.90 1.91 1.91 1.95 1.95 1.89 1.89 Rio Grande do Norte Rio Grande do Norte 0.83 0.83 0.80 0.80 0.80 0.80 0.83 0.83 0.87 0.87 0.86 0.86 Paraíba Paraíba 0.84 0.84 0.83 0.83 0.77 0.77 0.79 0.79 0.84 0.84 0.83 0.83 Pernambuco Pernambuco 2.39 2.39 2.31 2.31 2.27 2.27 2.32 2.32 2.34 2.34 2.34 2.34 Alagoas Alagoas 0.66 0.66 0.66 0.66 0.66 0.66 0.66 0.66 0.66 0.66 0.67 0.67 Sergipe Sergipe 0.64 0.64 0.64 0.64 0.63 0.63 0.63 0.63 0.64 0.64 0.63 0.63 Bahia Bahia 4.11 4.11 4.01 4.01 4.07 4.07 4.23 4.23 4.07 4.07 4.12 4.12 SUDESTE SUDESTE 56.68 56.68 55.75 55.75 55.83 55.83 56.53 56.53 56.79 56.79 56.41 56.41 Minas Gerais Minas Gerais 8.65 8.65 8.75 8.75 9.13 9.13 8.97 8.97 9.06 9.06 9.07 9.07 Espírito Santo Espírito Santo 1.81 1.81 1.83 1.83 2.07 2.07 2.20 2.20 2.23 2.23 2.27 2.27 Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 46 Espírito Santo Espírito Santo 1.81 1.81 1.83 1.83 2.07 2.07 2.20 2.20 2.23 2.23 2.27 2.27 Rio de Janeiro Rio de Janeiro 11.60 11.60 11.06 11.06 11.48 11.48 11.50 11.50 11.62 11.62 11.15 11.15 São Paulo São Paulo 34.63 34.63 34.11 34.11 33.14 33.14 33.86 33.86 33.87 33.87 33.92 33.92 SUL SUL 16.89 16.89 17.70 17.70 17.39 17.39 16.59 16.59 16.32 16.32 16.64 16.64 Paraná Paraná 5.98 5.98 6.44 6.44 6.31 6.31 5.90 5.90 5.77 5.77 6.07 6.07 Santa Catarina Santa Catarina 3.77 3.77 3.93 3.93 3.99 3.99 3.97 3.97 3.93 3.93 3.93 3.93 Rio Grande do Sul Rio Grande do Sul 7.14 7.14 7.33 7.33 7.10 7.10 6.72 6.72 6.62 6.62 6.64 6.64 CENTRO CENTRO-
  • OESTE
OESTE 8.77 8.77 9.01 9.01 9.11 9.11 8.86 8.86 8.71 8.71 8.87 8.87 Mato Grosso do Sul Mato Grosso do Sul 1.03 1.03 1.13 1.13 1.09 1.09 1.01 1.01 1.03 1.03 1.06 1.06 Mato Grosso Mato Grosso 1.42 1.42 1.64 1.64 1.90 1.90 1.74 1.74 1.49 1.49 1.60 1.60 Goiás Goiás 2.53 2.53 2.52 2.52 2.47 2.47 2.35 2.35 2.41 2.41 2.45 2.45 Distrito Federal Distrito Federal 3.80 3.80 3.71 3.71 3.64 3.64 3.75 3.75 3.78 3.78 3.76 3.76 BRASIL BRASIL 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00
slide-47
SLIDE 47

Broadband Price Estimation Broadband Price Estimation

  • No

No reliable reliable, , disaggregated disaggregated with with historical historical series series data available in Brazil data available in Brazil

  • The

The few data few data available were scattered from available were scattered from

  • The

The few data few data available were scattered from available were scattered from several sources, mainly from several sources, mainly from sporadic sporadic surveys surveys performed by telecom equipment performed by telecom equipment manufacturers, consultant companies or manufacturers, consultant companies or

  • ther organizations.
  • ther organizations.
Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 47
  • For academic purposes is desirable to count

For academic purposes is desirable to count

  • n official and more reliable data.
  • n official and more reliable data.
  • Anatel

Anatel started to collect prices data in 2011 started to collect prices data in 2011

  • nly
  • nly
slide-48
SLIDE 48
  • Alternative :

Alternative : estimate prices from survey estimate prices from survey data done by data done by CETIC.BR CETIC.BR , which is part of , which is part of

Broadband Price Estimation Broadband Price Estimation

data done by data done by CETIC.BR CETIC.BR , which is part of , which is part of the the CGI.br CGI.br (internet management (internet management committee) committee)

  • Annual surveys from CETIC.br ask :

Annual surveys from CETIC.br ask : “ “MAXIMUM DECLARED VALLUE BY USERS MAXIMUM DECLARED VALLUE BY USERS

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 48

“ “MAXIMUM DECLARED VALLUE BY USERS MAXIMUM DECLARED VALLUE BY USERS WILLING TO SUBSCRIBE INTERNET WILLING TO SUBSCRIBE INTERNET ACCESS SERVICE”. ACCESS SERVICE”.

slide-49
SLIDE 49

Broadband Price Estimation Broadband Price Estimation

  • Price x Penetration Model Adjusted:

Price x Penetration Model Adjusted:

Y = A . e Y = A . e-

  • B.X

B.X , where:

, where: Y: Y: number of broadband accesses per 1000 number of broadband accesses per 1000

  • Y:

Y: number of broadband accesses per 1000 number of broadband accesses per 1000 inhabitants. inhabitants.

  • X:

X: maximum declared price to subscribe the maximum declared price to subscribe the service. service.

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 49
  • A :

A : saturation. Penetration value when price

  • saturation. Penetration value when price

is zero. is zero.

  • B :

B : decay factor. decay factor.

slide-50
SLIDE 50

Broadband Price Estimation Broadband Price Estimation

  • Approximation of the broadband price

Approximation of the broadband price as a function of penetration : as a function of penetration :

( )

  • = −
Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 50
slide-51
SLIDE 51

Broadband Price Estimation Broadband Price Estimation

  • Same procedure for

Same procedure for 2005 2005 to to 2008 2008. .

$
  • '!
  • %

$ <

  • "
# $ %
  • &
  • <
< (%%?#")")**+#+* ,(?)%#$ Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 51
  • +
  • $
+ +$
  • $
"
  • !
,.!*
  • &
'!
slide-52
SLIDE 52

Broadband Price Estimation Broadband Price Estimation

  • For

For 2000 to 2004 2000 to 2004 used used 2005 2005 curve due lack of curve due lack of survey data for these years. survey data for these years.

  • The curves show a trend of being shifted as

The curves show a trend of being shifted as

  • The curves show a trend of being shifted as

The curves show a trend of being shifted as the times goes by and broadband prices the times goes by and broadband prices decrease. decrease.

  • Despite its limitations, it allowed to have a

Despite its limitations, it allowed to have a glimpse of the effect of the broadband price glimpse of the effect of the broadband price

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 52

glimpse of the effect of the broadband price glimpse of the effect of the broadband price

  • n the demand for the service.
  • n the demand for the service.
  • Other studies used the same approach

Other studies used the same approach ( (Coutinho Coutinho, Acorn , Acorn-

  • Redecom

Redecom 2010 2010). ).

slide-53
SLIDE 53

Models Used Models Used

  • Analysis without broadband prices:

Analysis without broadband prices: initial initial analysis because lack of analysis because lack of reliable reliable broadband broadband price data. price data.

  • Analysis with estimated broadband prices:

Analysis with estimated broadband prices: because the because the impact impact of price in the demand /

  • f price in the demand /

supply of the service, it was necessary to find supply of the service, it was necessary to find a way to estimate approximated prices using a way to estimate approximated prices using

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 53

a way to estimate approximated prices using a way to estimate approximated prices using survey data from survey data from CETIC (CGI). CETIC (CGI).

  • Prices estimated from maximum declared

Prices estimated from maximum declared values of users willing to subscribe internet values of users willing to subscribe internet access service. access service.

slide-54
SLIDE 54

Models Used Models Used

  • Analysis without

Analysis without education education and and demographic demographic density variables: density variables:

  • Excluded variables in the third type of model

Excluded variables in the third type of model because did not present the expected results: because did not present the expected results: because did not present the expected results: because did not present the expected results:

  • When including

When including price price, both , both education education and and demographic density demographic density had negative impact on had negative impact on broadband density. broadband density.

  • Possible cause:

Possible cause: price price has bigger impact when has bigger impact when compared to other variables compared to other variables (price elasticity in (price elasticity in

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 54

compared to other variables compared to other variables (price elasticity in (price elasticity in Brazil from Brazil from -

  • 1

1 to to -

  • 3,36

3,36, , Ávila (2008) Ávila (2008)) ), reducing the , reducing the importance of importance of education education and and demographic density demographic density in the model. in the model.

slide-55
SLIDE 55

Simultaneous Equations Model: Simultaneous Equations Model: Main Results Main Results

  • Each

Each 1 p.p. 1 p.p. increase in the increase in the broadband penetration is related to broadband penetration is related to broadband penetration is related to broadband penetration is related to a a GDP growth GDP growth in Brazil between in Brazil between 0.038 0.038 to to 0,18 0,18 p.p. p.p.

  • Each

Each1 p.p. 1 p.p. increase in the broadband increase in the broadband

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 55
  • Each

Each1 p.p. 1 p.p. increase in the broadband increase in the broadband penetration is related to a penetration is related to a GDP per GDP per Capita growth Capita growth in Brazil between in Brazil between 0.196 0.196 to to 0,362 0,362 p.p. p.p.

slide-56
SLIDE 56

Simultaneous Equations Model: Simultaneous Equations Model: Main Results Main Results

  • However it was not possible to establish

However it was not possible to establish causality relation between increase of causality relation between increase of causality relation between increase of causality relation between increase of the Broadband Penetration and the Broadband Penetration and economic growth. economic growth.

  • For that it would be necessary other

For that it would be necessary other approach using long term relationship approach using long term relationship

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 56

approach using long term relationship approach using long term relationship (long term time series and (long term time series and Granger Granger causality test causality test) )

slide-57
SLIDE 57

Comparison with Other Studies Comparison with Other Studies

Values of Regression Coefficients Values of Regression Coefficients Earlier Studies Earlier Studies This Study This Study

Dependent Dependent Variable Variable Explaining Explaining Variables Variables World Bank World Bank ( (Qiang Qiang et. al,

  • et. al,

2009, p. 44) 2009, p. 44) Koutroumpis (2009, Koutroumpis (2009, p.478) p.478) Simultaneous Equations Simultaneous Equations Simultaneous Simultaneous Equations Equations 2009, p. 44) 2009, p. 44) Simultaneous Equations Simultaneous Equations

GDP GDP

Broadband Broadband Penetration Penetration (number of (number of accesses accesses per per

0,012, 0,012, 0.023,

0.023, 0.025 and 0.025 and 0.204

0.204 0.038, 0.038, 0.097,

0.097, 0.109 , 0.104, 0.109 , 0.104, 0.133 and 0.133 and 0.180

0.180

0.138 0.138

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 57

per per inhabitants) inhabitants)

GDP per GDP per Capita Capita

0.138 0.138

(developing (developing countries countries) )

0.121 0.121

(developed (developed countries countries) )

0.196 , 0.196 , 0.203 ,

0.203 , 0.238, 0.260, 0.238, 0.260, 0.321 0.321 and and 0.362

0.362

slide-58
SLIDE 58

Results: Comments Results: Comments

  • Other possible reasons:

Other possible reasons:

  • Differences in in the historical moment when

Differences in in the historical moment when broadband networks became available: broadband networks became available:

  • Developed countries:

Developed countries: arrived when they already had arrived when they already had

  • Developed countries:

Developed countries: arrived when they already had arrived when they already had a good telecom infrastructure with well developed a good telecom infrastructure with well developed fixed line and mobile phone networks: fixed line and mobile phone networks:

  • So broadband economic impact was diluted

So broadband economic impact was diluted as just one as just one more development increase factor, because those more development increase factor, because those countries already had reached high development levels. countries already had reached high development levels.

  • Developing countries:

Developing countries: broadband networks became broadband networks became

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 58
  • Developing countries:

Developing countries: broadband networks became broadband networks became an important instrument to boost development an important instrument to boost development (“catch up” effect): (“catch up” effect):

  • Helping develop education areas and others.

Helping develop education areas and others.

slide-59
SLIDE 59

Key Message Key Message

  • Despite all limitations of this study:

Despite all limitations of this study:

  • It was a

It was a first attempt first attempt to make this kind of study to make this kind of study for for Brazil

  • Brazil. Expectation that it would

. Expectation that it would “open the “open the gates” gates” to further and to further and more precise more precise studies, studies, gates” gates” to further and to further and more precise more precise studies, studies,

  • vercoming the limitations
  • vercoming the limitations and errors of this one.

and errors of this one.

  • It is important that

It is important that each country make an effort each country make an effort to develop its own broadband impact evaluation to develop its own broadband impact evaluation studies studies, instead of using estimations made for , instead of using estimations made for

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 59

studies studies, instead of using estimations made for , instead of using estimations made for

  • ther countries with different realities.
  • ther countries with different realities.
  • You have to know your own reality in order to

You have to know your own reality in order to develop adequate solutions for the problems that develop adequate solutions for the problems that need to be addressed. need to be addressed.

slide-60
SLIDE 60

Regression Results: Comments Regression Results: Comments

  • It may occur R

It may occur R2

2 < 0 in models without a constant

< 0 in models without a constant (intercept) and which are unable to explain properly the (intercept) and which are unable to explain properly the dependent variable. dependent variable.

  • −

∑

  • : sample value

: sample value i i of explaining variable

  • f explaining variable y

y. .

  • : estimated values by the regression model.

: estimated values by the regression model.

  • =

=

− = − = − −

∑ ∑

  • Broadband Economic Impact in Brazil: a
Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 60
  • : estimated values by the regression model.

: estimated values by the regression model.

  • : average of sample values.

: average of sample values.

slide-61
SLIDE 61

Regression Results: Comments Regression Results: Comments

  • : squared sum of errors between estimated

: squared sum of errors between estimated and real sample values. and real sample values.

  • : squared sum of differences between each

: squared sum of differences between each sample value and the average value of all samples. sample value and the average value of all samples.

  • =

= −

∑

  • =

= −

∑

sample value and the average value of all samples. sample value and the average value of all samples.

  • When SSR > SST, the model errors are bigger than the

When SSR > SST, the model errors are bigger than the errors obtained if the model was represented just by the errors obtained if the model was represented just by the average value of the samples. average value of the samples.

  • It may happen in the lack of a constant (intercept) in the

It may happen in the lack of a constant (intercept) in the

  • =
Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 61

It may happen in the lack of a constant (intercept) in the It may happen in the lack of a constant (intercept) in the equation model and when the model itself is insufficient equation model and when the model itself is insufficient to explain the dependent variable behavior. to explain the dependent variable behavior.

  • So, when R

So, when R2

2 <

< < < < < < < 0, the 0, the R R2

2 loses its characteristic of “quality”

loses its characteristic of “quality” indicator of the model. indicator of the model.

slide-62
SLIDE 62 !
  • "
#
  • !

Broadband Infrastructure Production Function Broadband Infrastructure Production Function

( )

  • ε

β β + ⋅ + =        

−
  • Difficulty to

Difficulty to match the model match the model

# $ " " % & & ' $ " " " # " % " $ # & ( " # % ) $
  • "
# % ) $

match the model match the model with real data. with real data.

  • Lack of intercept.

Lack of intercept.

Broadband Economic Impact in Brazil: a Simultaneous Equations Analysis Hildebrando R. Macedo and Alexandre X. Y. Carvalho 5th ACORN-REDECOM Conference Lima,Peru, May 19-20th, 2010 62 " # % ) $ $ " # % ) $ # % $ # % " # % ) $ & (
  • That lead to

That lead to R2

2 < 0

< 0