Big Brother 19 | 3DS (69) | Towa no Quon

The importance of ICT data to measure its social and economic impact


Save this PDF as:
 WORD  PNG  TXT  JPG

Tamaño: px
Comenzar la demostración a partir de la página:

Download "The importance of ICT data to measure its social and economic impact"

Transcripción

1 The importance of ICT data to measure its social and economic impact Dr. Raúl L. Katz, Adjunct Professor, Division of Finance and Economics, and Director, Business Strategy Research, Columbia Institute of Teleinformation Workshop INDICADORES DE TIC PARA AVALIAÇÃO DO DESENVOLVIMENTO ECONÔMICO E SOCIAL Brasilia, November 8, 2011 MINISTÉRIO DAS COMUNICAÇÕES SECRETARIA EXECUTIVA

2 Contents ICT Evidence- based policy making and the need of data Measuring broadband impact: data and analy>cal problems A case study of economic impact measurement of a na>onal broadband plan The next fron>er in ICT impact analysis: digi>za>on 2

3 Public policy decisions are, in many cases, not made in terms of social and economic impact analysis OCen >mes, countries and policy makers tend to follow examples of their peers according to three paferns Geographic proximity: diffusion is driven by similari>es in condi>ons among states ( copy your neighbor ), frequent interac>on of officials of neighboring countries Lateral diffusion: policy imita>on of countries sharing common cultural and socio- economic factors Hierarchical diffusion: models are developed in the most advanced or large countries, and then adopted by successively less advanced or smaller na>ons Herd behavior: the decision is determined not only by a policy maker preferences and informa>on, but also by the behavior of other countries The signals of certain countries shape the behavior and beliefs of policy makers, groups in other countries When these signals lead to a convergence of large group of countries, we witness herd behavior Convergence: Over >me, however, spectrum alloca>on models grow more alike in terms of increasing similari>es Harmoniza>on through interna>onal agreements or supra- na>onal regula>ons Coercive imposi>on of models 3

4 Evidence-based approaches are considered as tools to bring some rigor to policy making Evidence- based policy is public policy formulated on the basis of rigorously established objec>ve evidence Evidence- based policy approaches can be applied to telecommunicac>ons policy making Tests why a certain policy recommenda>on is more cost- effec>ve and what the impact of the policy will be if it is successful Includes a counterfactual: what will occur if the policy is not implemented Incorporates a quan>ta>ve measurement of the impact of the policy Examines both direct and indirect effects that occur because of the policy ( Law of unintended consequences ) Separates the uncertain>es and controls for other influences outside of the policy that may have an effect on the result Impact assessment should be able to be tested by a third party The overall methodology can be fit under the category of cost- benefit analysis used to es>mate a net payoff if the policy is implemented 4

5 Why is data important in measuring ICT economic impact? An imperfectly informed policy maker is constrained in his choice of the most appropriate approaches to overseeing the ICT sector Governments ocen cannot formulate correct public ICT policies Under condi>ons of limited data on ICT impact, governments tend to imitate/copy the policy and regulatory models of developed countries Common cases of mis- informa>on due to the lack of data The Solow paradox: is broadband affec>ng produc>vity? Size of the broadband supply gap in the United States: 20 million households or 10 million households? Reasons explaining the broadband demand gap: why do some countries have low broadband penetra>on? Broadband and job crea>on: does broadband contribute to employment? 5

6 ICT data is the required basis for conducting rigorous policy costbenefit analysis Cost- benefit analysis feeding regulatory decisions requires accurate data Limited data introduces considerable noise in the analy>cal process Considering the amount of investment in ICT, and their economic impact, the amount of data and analysis leading to decisions is typically sub- op>mal In our view, given the lifle data some>mes policy makers have about economic impact of ICT, this is not an issue of how to op>mize regula>on under imperfect informa>on, but how do we gain a befer understanding of these effects 6

7 Lack of ICT data fosters policy imitation One or more incen>ves for imi>a>ve behavior can be iden>fied Reduc>on of uncertainty: the more countries adopt a policy, the higher the value of example Promote reputa>on Cost of informa,on: herding allows to short- cut the analy,cal work or use of rigorous data analysis leading to the formula,on of a regulatory model Learning from opinion leaders: leaders that are outside the country of decision maker increase legi>macy of decision 7

8 Lack of impact data results in defining policy targets without underlying rigorous analysis: example of broadband speeds TARGETS OF COVERAGE AND DOWNLOAD SPEED FOR NATIONAL BROADBAND PLANS COUNTRY Coverage Targets (as a percent of households) Speed Targets (as a percent of households) United States 100 % (2012) 4 Mbps (100%) (2012) 50 Mbps Germany 100 % (2014) 1 Mbps (100%) (2014) 50 Mbps (75%) (2014) Singapore 100 % (2012) 100 Mbps (95%) (2012) Australia 100 % (2012) 12 Mbps (100%) (2012) United Kingdom 100 % (2012) 2 Mbps (100%) (2012) Malaysia 75 % (2010) (33%) Mbps (42%) 1.5 Mbps Brazil 50% hogares urbanos y 25% hogares rurales 75% ( kbps) European Union 100 % (2013) 30 Mbps (100%) (2020) 100 Mbps (50%) (2020) Sources: National Broadband Plans 8

9 Coverage and service targets need to be defined on the basis of rigorous analysis of level of investment and social and economic returns Existing infrastructure Coverage and service targets Infrastructure gap Modernization of existing lines to deliver target service levels Deployment of new lines to achieve coverage targets Required technology by zone Economic benefits Construction effect (multipliers) Contribution to GDP growth Job creation Spill over effects (innovation, new business creation) Consumer surplus Access to public services and information Savings in transport time Health and education services Cost per line Total investment requirement RETURN ON INVESTMENT 9

10 Contents ICT Evidence- based policy making and the need of data Measuring broadband impact: data and analy>cal problems A case study of economic impact measurement of a na>onal broadband plan The next fron>er in ICT impact analysis: digi>za>on 10

11 ICT Data gathering should be guided by an understanding of impact processes: the Broadband example BROADBAND ECONOMIC IMPACT Broadband deployment Direct benefits Investment in infrastructure deployment Residential penetration Business penetration Household income Total factor productivity Consumer surplus Contribution to GDP growth and employment 11

12 Research on the economic impact of broadband focuses on three areas Impact of externalities BROADBAND ECONOMIC IMPACT Creation of consumer surplus Residential penetration Consumer surplus Broadband deployment Direct benefits Investment in infrastructure deployment Business penetration Household income Total factor productivity Contribution to GDP growth and employment Contribution to employment and output of broadband deployment 12

13 Infrastructure effect: Three types of network construction effects exist EFFECT DESCRIPTION EMPLOYMENT EXAMPLES Direct jobs and output Employment and economic production generated in the short term in the course of deployment of network facilities Indirect jobs and output Employment and production generated by indirect spending (or businesses buying and selling to each other in support of direct spending) Induced jobs and output Employment and production generated by household spending based on the income earned from the direct and indirect effects Telecommunications technicians Construction workers Civil and RF engineers Metal products workers Electrical equipment workers Professional Services Consumer durables Retail trade Consumer services 13

14 Infrastructure effect: Estimates from several countries indicate that network construction effects and multipliers are significant NETWORK CONSTRUCTION EFFECTS OF BROADBAND COUNTRY RESEARCHER / INSTITUTION STIMULUS INVEST. (US$ million) NETWORK DEPLOYMENT JOBS ESTIMATE DIRECT INDIRECT INDUCED TOTAL TYPE I (*) MULTIPLIERS TYPE II (**) UNITED STATES SWITZERLAND GERMANY UNITED KINGDOM AUSTRALIA Katz (Columbia) Atkinson (ITIF) Katz (Columbia) Katz (Columbia) Liebenau (LSE) Government $ 6,390 37,300 31,000 59, , $ 10,000 63, , , ~$ 10,000 ~80,000 ~30,000 N.A. ~110, N.A. $ 47, , , , , $ 7,463 76, , , $ 31,340 ~200,000 Sources: Katz, R. and Suter, S. (2009). Estimating the economic impact of the US broadband stimulus plan, Columbia Institute for Tele-Information working paper; Katz, R., P. Zenhäusern, S. Suter, P. Mahler and S. Vaterlaus (2008). Economic Modeling of the Investment in FTTH in Switzerland, unpublished report; Libenau, J., Atkinson, R. (2009) The UK s digital road to recovery. LSE and ITIF; Australian government. Katz, R., S. Vaterlaus, P. Zenhäusern, S. Suter and P. Mahler (2009). The Impact of Broadband on Jobs and the German Economy; Columbia Institute for tele-information working paper (*) (Direct + indirect)/direct (**) (Direct + indirect + induced)/direct 14

15 Infrastructure effect: data to support a quantification of investment (input) and the I-O matrices are required to estimate this DATA UTILIZED IN INPUT-OUTPUT STUDIES DATA REMARKS AVAILABILITY RATIONALE Investment in broadband program Breakdown of investment by sectors (i.e. manufacturing of electronic equipment, construction, telecommunication) Input-output-table Requires reformatting to estimate employment effects To be calculated based on benchmarks To be supplied by government statistical units and/ or Central Banks This is the investment input that will trigger growth in output and/or jobs Required to understand inter-sectoral relationships # of employed persons In the same classification as the input-output-table To be supplied by government statistical units Required to calculate employment effects 15

16 Infrastructure effect: Input-output model in general terms The input- output model of economics uses a matrix representa>on of a na>on's economy. It is designed to predict the effect of addi>onal demand in one industry (e. g. telecommunica>ons) on others. The input- output analysis considers inter- industry rela>ons in an economy. It depicts how the output of one industry goes to another industry where it serves as an input, and thereby makes one industry dependent on another both as customer of output and as supplier of inputs. An input- output model is a specific formula>on of input- output analysis. The mathema>cs of input- output economics is straighborward, but the data requirements are enormous because the expenditures and revenues of each branch of economic ac>vity has to be represented. 16

17 Infrastructure effect: The input/output-matrix assesses the value added of additional demand based on a multiplier Input side Goods/Sector Output side (use side) Sector Input- Output table (each column of the input- output matrix reports the monetary value of an industry's inputs and each row represents the value of an industry's output). Inputs + Value added = Gross produc>on + Imports Volume of goods = Inputs + End demand Households State Investments Exports Volume of Goods = - = Imports Gross domes>c product With input- output tables it is possible to calculate the effect of addi>onal investment in telecommunica>ons The result is a mul>plier indica>ng how much addi>onal value added will be generated due to the fiber deployment. What measure Mul>pliers? Mul>pliers measure the total addi>onal produc>on in the economy due to a unit increase in demand in a specific sector (e.g. construc>on). 17

18 Infrastructure effect: The multiplier model is built based on the following data sources Input- output- matrix Data available and processed Hours worked and employment data Data available and processed Capex data from operators Data available and processed 18

19 Infrastructure effect: Example: Total investment of fiber rollout is split between sectors * Numbers according to operator s cost alloca,on; we need this split to assign the different industries in the input/output- matrix 19

20 Infrastructure effect: Total additional value added through fiber rollout * The fiber rollout is crea>ng a total addi>onal produc>on of mchf About mchf are imported goods and mchf are produced within Switzerland The addi>onal produc>on in Switzerland generates about mchf intermediate inputs and mchf value added Investment in fiber rollout Total additional production Domestic additional production Additional value added 13'196 22'791 19'565 9'853 Imported goods Intermediate inputs 3'226 9'712 * All rela,ons are a result of the input- output- matrix 20

21 Infrastructure effect: Direct, indirect effects and the multiplier The difference between total addi>onal value added in the economy and the addi>onal value added in the inves>ng sectors is the indirect effect. The rela>onship between the total effect in the economy and the direct effect defines the mul>plier. Direct, indirect effects and mul,pliers Value Added Direct Effect (value added to construc>on, telecom etc.) 6'762 m CHF Indirect Effect (value added to other sectors of the economy ) 3'091 m CHF Total Effect 9'853 m CHF Mul>plier 1.46 Employment Direct Effect 82,800 employees Indirect Effect 31,400 employees Total Effect 114,100 employees Mul>plier 1.38 Work Volume Direct Effect m hours Indirect Effect 53.3 m hours Total Effect m hours Mul>plier

22 Network effects: The externalities derived from broadband are significantly higher EFFECT DESCRIPTION EMPLOYMENT EXAMPLES Productivity Improvement of productivity as a result of the adoption of more efficient business processes enabled by broadband Marketing of excess inventories Optimization of supply chains Innovation Acceleration of innovation resulting from the introduction of new broadbandenabled applications and services New applications and services (telemedicine, Internet search, e- commerce, online education, VOD and social networking) Value chain recomposition Attract employment from other regions as a result of the ability to process information and provide services remotely New forms of commerce and financial intermediation Outsourcing of services Virtual call centers Core economic development clusters 22

23 Network effects: Research to date confirms the contribution to GDP growth but the amount of impact varies widely Contribution to GDP growth of 10% increase in broadband penetration RESEARCH EVIDENCE OF BROADBAND IMPACT ON GDP GROWTH RESEARCHER /INSTITUTION Low-Med. Income High Income OECD-High OECD-Low UK US France Germany OECD-High OECD-Med OECD-Low Germany-High Germany-Low L. America Waver- Kou- Qiang Czernich man Katz troumpis World Bank U. Munich LECG Imperial C. U. Columbia Brazil Chile India Malaysia 23

24 Network effects: However, these estimates are consistent with growing evidence of increasing returns to broadband penetration INCREASING BROADBAND IMPACT ON GDP GROWTH Country Average % Impact of BB on growth Cluster average impact on growth High penetration Denmark, Norway, Netherlands, Sweden, Switzerland Average contribution to GDP growth: Low penetration Greece, Portugal, Italy, New Zealand, Austria, Hungary, Spain, Ireland Average contribution to GDP growth: % 14% 16% 17% 21% 22% 22% 24% Broadband Penetration (2007) Source: adapted from Koutroumpis (2009) 25% 30% 33% 0 Medium penetration Germany, France, Japan, Belgium, UK, Australia, US, Canada, Luxemburg Average contribution to GDP growth:

25 Network effects: The contribution of broadband externalities to employment comprises three simultaneous effects + e-business impact on firm productivity + Macroeconomic productivity - Impact on employment - Incremental broadband penetration + Enhanced innovation + + Impact on employment Impact on employment + + -/+ Outsourcing of services + Displacement to service sector Note: This causality chain was adapted from a model originally developed by Fornefeld et al., 2008 in a report for the European Commission 25

26 Network effects: These effects result in different output and employment impact depending on broadband penetration High Broadband Penetration Regions Low Broadband Penetration Regions HI GDP HI GDP Economic Impact LO Employment Economic Impact LO Employment Increase in BB penetration T+1 T+2 T+3 T+4 Increase in BB penetration T+1 T+2 T+3 T+4 High economic growth initially, diminishing over time ( supply shock effect) New Economic Growth (innovation, new services) High stable economic growth ( catch up effect) Capital/labor substitution limits employment growth ( productivity effect ) 26

27 Network effects: an increase in broadband penetration of 10% will contribute to Latin American GDP growth in 0.16% Promedio Crec. PIB = β1*pib Capita + β2*(inversión/pib) + β3*nivel Educación Terciaria + β4*( de Banda Ancha)+β5* Crecimiento de la población + β5*indice d e Globalización Coeficiente Error estándar T-statistic P>[t] Penetracion de banda ancha PIB Capita Inversion/PIB Variables de control Nivel Educación Terciaria Crecimiento de Población Índice de Globalización Constante Número de observaciones = 49 F(4, 14) = 7.18 Prob> F = R2 = Root MSE = INTERPRETACIÓN La diferencia en el impacto con respecto a la estimación del Banco Mundial podría deberse a que el este último incluye a países desarrollados que han alcanzado o sobrepasado el umbral de masa crítica en la penetración de banda ancha El valor explicativo de la inversión de capital no es significativo porque ha transcurrido poco tiempo entre la inversión y el crecimiento El uso de educación terciaria se explica para medir las diferencias en el capital humano 27

28 Network effects: data requirements for econometric analysis include both ICT and economic DEPENDENT VARIABLES Annual or quarterly rate of change of GDP Annual or quarterly rate of change of employment Annual or quarterly rate of change of unemployment Annual number of SME s INDEPENDENT VARIABLES Annual or quarterly rate of change of broadband penetration DATA UTILIZED IN ECONOMETRIC STUDIES CONTROL VARIABLES GDP at starting time of period Level of education: Percent of population with tertiary degrees; Illiteracy rate; Years of schooling; participation rate in secondary school Regional Investment as percentage of regional GDP Percent of households with electricity or running water Number of projects and added value of construction projects financed by the state Number of hospitals per inhabitant; number of beds in hospitals per pop. Access to financial services: Number of banking offices and bank credit p.c. Industry concentration: Contribution of financial services, commerce and manufacturing sectors to regional GDP Importance of tourism in the region (number of domestic tourism trips) Cost index for interstate trade costs Cost to create new business Regional Gini Coefficient Percentage of people living in urban centers Total road length per hundred sq. Km by area; Road development index Population growth rate Globalization Index; Globalization Index per region 28

29 Consumer surplus: utility gain of broadband that can be purchased at a price lower than what consumers are willing to pay Benefits that broadband technology (DSL, cable modem, Fttx) yields in relation to dial-up Benefits derived from increased broadband penetration (access to new services and information) Consumer surplus = Willingness to Pay - Price of service Price declines as a result of competition, technology trends, and productivity gains 29

30 Consumer surplus: this approach can be used either historically or forward-looking HISTORICAL STUDIES FORWARD-LOOKING Greenstein and McDevitt (2009) analysis of broadband bonus in the United States Crandall and Jackson (2001) analysis of consumer surplus generated by broadband adoption in the United States Greenstein and McDevitt (2010) comparative analysis of consumer value generated by broadband diffusion in China, Mexico, Spain, Canada and the United Kingdom: scale of value creation is proportionate to deployment Katz et al. (2008) comparative analysis of consumer surplus to be generated in Switzerland by alternative fiber deployment strategies: consumer surplus is maximized in competitive models that promote infrastructure based competition 30

31 Advantages and disadvantages of each methodology Input-output analysis Econometric analysis ADVANTAGES Easy to communicate Based on proven interlinks Ability to link projections of broadband penetration to growth DISADVANTAGES Static models Numerous caveats on induced effects Dependent on up to date I-O matrices Length of data sets Data disaggregation Consumer Surplus Useful to calculate economic impact not captured GDP numbers Results are quite valid in the short run because the analysis assumes stable demand Analysis could be conservative because it excludes gains to early consumers, shifts in demand due to GDP growth, falling prices of PCs, etc. 31

32 To sum up, data requirements for assessing broadband economic impact range from aggregate macro to the micro-data data Disaggregated data for ICT, broadband and economic indicators Increase the number of observa>ons Deal with fixed effects Quarterly data Gain befer understanding of >ming of effects Ability to conduct analysis for monitoring of effects Data on broadband download speeds Understand return to broadband speed Frame investment and broadband target requirements Data on wireless internet access Capture a growing trend and shic in broadband plaborm Broadband coverage data Understand uptake trends Gain more evidence to support formula>on of demand gap policies Variables to control income endogeneity (prices, compe>>on, government, investment) Gain befer understanding of causality 32

33 Contents ICT Evidence- based policy making and the need of data Measuring broadband impact: data and analy>cal problems A case study of economic impact measurement of a na>onal broadband plan The next fron>er in ICT impact analysis: digi>za>on 33

34 Este estudio tiene como objetivo estimar el impacto economico del plan de banda ancha de un pais Para la evolución, alineamiento, coordinación y complementariedad de la Estrategia de banda ancha, en el marco de las polí>cas del Plan Nacional de Desarrollo, se hace necesario contar con una asesoría para la evaluación del impacto socio- económico del Plan de banda ancha El estudio requerido se enfocará en tres áreas de impacto de la banda ancha como tecnología de base para la entrega de servicios de gobierno en línea: La contribución al crecimiento del PIB El impacto en la creación de fuentes de trabajo El impacto en la reducción de la pobreza 34

35 El analisis esta basado en la construccion de modelos econometricos que estiman el impacto economico a la fecha de la banda ancha en el pais Las es>maciones están basadas en un análisis del impacto histórico de las variables independientes (por ejemplo, la penetración de banda ancha), que se ven iden>ficadas como metas del plan de banda ancha Una vez estas determinadas se construyen modelos econométricos que determinan cuál ha sido el impacto de dichas variables en las tres áreas de impacto en el país Para ello, se compila información historica desagregada a nivel estatal Con ello se construye un panel de datos que incluye la penetración de banda ancha por hogar o por habitantes, la evolución del PIB departamental, la tasa de empleo, así también como variables económicas, de capital humano, y de infraestructura que actúan como control de efectos que no están ligados a la banda ancha (por ejemplo, inversión bruta de capital fijo, años de educación, etc.) 35

36 Las bases de datos utilizadas son aquellas residentes en las instituciones dedicadas al planeamiento economico Hay que destacar que en todas las especificaciones se esta trabajando con datos a nivel estatal para los años 2007, 2008, 2009 y 2010 Se nota que también se cuenta con la información para el año 2006, que se u>liza para obtener el crecimiento de las diversas variables al año 2007) Los estados analizados son los existentes al año 2011 Pese a que se disponía de información económica y social a nivel estatal desde el año 2001, la falta de disponibilidad de la can>dad de conexiones de banda ancha por estado no permite su u>lización antes del

37 Para estudiar el efecto del crecimiento de la banda ancha, se utiliza el crecimiento anual a nivel estatal Se contabilizan los accesos con velocidad de descarga igual o superior a 256 kbps (Definición de Banda Ancha u>lizada por OECD) según información provista por el ministerio de comunicaciones Según la situación de los estados en relación a la tasa de penetración (u>lizando el número de conexiones previamente mencionado y la población departamental) al año 2010, se divide a la muestra en estados con alta tasa de penetración y estados con baja tasa de penetración 37

38 El analisis de impacto es realizado para dos sub-muestras, como se indica en el cuadro para el año 2010 ESTADOS ALTA PENETRACIÓN ESTADO ACCESOS B.A./ POBLACIÓN ESTADOS BAJA PENETRACIÓN ESTADO ACCESOS B.A./ POBLACIÓN Fuente: Elaboración de Telecom Advisory Services 38

39 Una vez construidos los paneles, se desarrollan los modelos econometricos de impacto Los modelos econométricos de impacto están construidos en base a los datos del país Los modelos econométricos están basados en los estudios ya realizados en el contexto de los trabajos realizados para la CEPAL, la UIT, y Alemania Katz, R. (2010). "La contribución de la banda ancha al desarrollo económico", V. Jordán, W. Peres y H. Galperin (eds.), Acelerando la revolución digital: banda ancha para América La8na y el Caribe, Comisión Económica para América La>na y el Caribe (CEPAL) / Diálogo Regional sobre Sociedad de la Información (DIRSI), San>ago de Chile Katz, R. L., Vaterlaus, S., Zenhäusern, P. & Suter, S. (2010). The Impact of Broadband on Jobs and the German Economy. Intereconomics, 45 (1), Katz, R. (2011). The impact of broadband on the economy: Research to date and policy issues, Interna>onal Telecommunica>on Union, Trends in Telecommunica8on Reform Geneva Sin embargo, los modelos están especificados en función de las variables de la realidad del país 39

40 A partir de los modelos, se determina el impacto a generarse a partir del cumplimiento de las metas del plan de banda ancha ABORDAJE METODOLOGICO COMPILACION DE SERIES HISTORICAS CONSTRUCCION DE MODELOS ECONOMETRICOS PROYECCION DEL IMPACTO DE PLAN DE BANDA ANCHA CRECIMIENTO DE BANDA ANCHA ( ) CRECIMIENTO DE BANDA ANCHA ( ) IMPACTO ECONOMICO ( ) PIB Empleo Ingreso del hogar IMPACTO ECONOMICO ( ) PIB Empleo Ingreso del hogar 40

41 El siguiente capitulo presenta los tres modelos de impacto economico historico de la banda ancha en el pais A Impacto del crecimiento de la penetración de la banda ancha en el incremento del producto interno bruto B Impacto del crecimiento de la banda ancha en el crecimiento promedio del ingreso de los hogares (como proxy de la disminucion de la pobreza) C Contribución de la banda ancha a la creación de fuentes de trabajo 41

42 A Primero, se estima el impacto de la banda ancha en el crecimiento del PIB, controlando por el PIB inicial y el crecimiento de la población CONTRIBUCION DE LA BANDA ANCHA EN EL CRECIMIENTO DEL PBI Crecimiento del PBI (%) Variable Dependiente: Crecimiento del PBI entre 2006 y 2010 Variables Independientes: Crecimiento accesos B.A., PBI 2003 y Crecimiento Población Total Baja Penet. Alta Penet. Crecimiento accesos B.A. (%) *** *** *** ( ) ( ) ( ) Crecimiento Población (%) * *** ( ) ( ) ( ) PBI 2003 (Miles de Millones de Pesos) * ( ) ( ) ( ) R^2 ajustado Número de Observaciones Nota: ***, ** y * indican significancia a un nivel del 1%, 10% y 15% El coeficiente señala que si las conexiones de banda ancha aumentan en un 10%, el PIB aumentaría en 0.037% 42

43 A El coeficiente señala que si las conexiones de banda ancha aumentan en un 10%, el PIB aumentaría en 0.037% Si bien el coeficiente puede parecer bajo (0.037), es interesante contextualizar el aumento promedio de la can>dad de conexiones, y del PIB en pesos constantes para el período bajo estudio EVOLUCIÓN MODELO CRECIMIENTO DEL PIB Promedio Variables por Departamento Fuente: Elaboración de Telecom Advisory Services, LLC Total B.Penet. A.Penet. Total B.Penet. A.Penet. Crecimiento PBI (%) Crecimiento Accesos B.A. (%) Crecimiento Población (%) El cuadro indica que para el total de las observaciones tenemos un aumento de puntos porcentuales en promedio de crecimiento anual de accesos de banda ancha, debido a fuertes aumentos de la tasa de penetración para la mayorías de los estados, triplicando su can>dad de abonados entre períodos Mul>plicando este valor por el efecto en el crecimiento del PIB, notamos que la variable bajo estudio contribuye en promedio 0.80 puntos porcentuales de crecimiento del PIB anualmente ( * ), lo que explica la mitad del aumento del PBI en estos años 43

44 A El modelo tambien indica que el impacto economico de la banda ancha tiende a aumentar conforme aumenta la penetracion Como se mencionó en la sección anterior, el analisis de impacto económico fue realizado de manera agregada para el país y dividiendo la muestra en departamentos cuya penetración de banda ancha excede el 2% y departamentos con una penetración es inferior a este parámetro Para los estados de baja penetración, la contribución de la banda ancha puede expresarse de la siguiente manera: si se aumenta la penetración en 10%, el PIB departamental crece en 0.039% Para los estados de alta penetración, si la penetración aumenta en un 10%, la contribución económica será de 0.044% 44

45 A Asimismo, al incluir variables de control por capital humano e inversion en capital fisico, se comprueba que el modelo es estable En el siguiente cuadro se incluyen variables de control por capital humano (años de educación promedio de la población en base a datos del Censo del 2005) EFECTO EN CRECIMIENTO DEL PBI, CONTROLANDO POR CAPITAL HUMANO Crecimiento del PBI (%), con control por capital humano Variable Dependiente: Crecimiento del PBI entre 2006 y 2010 Variables Independientes: Crecimiento accesos B.A., PBI 2003, Años de Educación y Crecimiento Población Total Baja Penet. Alta Penet. Crecimiento accesos B.A. (%) *** *** *** ( ) ( ) ( ) Crecimiento Población (%) ** *** ( ) ( ) ( ) Años de Educación ( ) ( ) ( ) PBI 2003 (Miles de Millones de Pesos) ( ) ( ) ( ) * R^2 ajustado Número de Observaciones Nota: ***, ** y * indican significancia a un nivel del 1%, 10% y 15% Fuente: Elaboración de Telecom Advisory Services 45

46 B El segundo modelo estima el impacto del crecimiento de banda ancha en el crecimiento del ingreso real por hogar EFECTO EN CRECIMIENTO DEL INGRESO REAL POR HOGAR Crecimiento del Ingreso Real Por Hogar (%), con control por capital humano Variable Dependiente: Crecimiento del Ingreso Real Por Hogar entre 2006 y 2010 Variables Independientes: Crecimiento accesos B.A., Crecimiento Población, Años de Educación, Producto del sector Minero y Viviendas NBI 2005 Total Baja Penet. Alta Penet. Crecimiento accesos B.A. (%) *** *** *** ( ) ( ) ( ) Crecimiento Población (%) *** *** * ( ) ( ) ( ) Años de Educación ** ( ) ( ) ( ) Producto del Sector Minero (%) *** *** ( ) ( ) Viviendas NBI 2005 (%) ** ( ) ( ) *** ( ) R^2 ajustado Número de Observaciones Nota: ***, ** y * indican significancia a un nivel del 1%, 10% y 15% Fuente: Elaboración de Telecom Advisory Services 46

47 C El tercer modelo estima el impacto del crecimiento de banda ancha en el crecimiento de la tasa de empleo por departamento Este modelo controla por el PIB de tres períodos previos y por el crecimiento de la población Fuente: Elaboración de Telecom Advisory Services EFECTO EN TASA DE EMPLEO Crecimiento de tasa de empleo (%) Variable Dependiente: Crecimiento de tasa de empleo entre 2006 y 2010 Variables Independientes: Crecimiento accesos B.A., PBI 2003 y Crecimiento Población Total Baja Penet. Alta Penet. Crecimiento accesos B.A. (%) ** ** ( ) ( ) ( ) Crecimiento Población (%) * ( ) ( ) ( ) PBI 2003 (Miles de Millones de Pesos) ( ) ( ) ( ) R^2 ajustado Número de Observaciones Nota: ***, ** y * indican significancia a un nivel del 1%, 10% y 15% 47

48 En resumen, los modelos permiten estimar el impacto economico de la banda ancha con suficiente robustez estadistica IMPACTO ECONOMICO DE LA BANDA ANCHA ( ) SI LA PENETRACION DE BANDA ANCHA AUMENTA EN 10 PUNTOS PORCCENTUALES IMPACTO EN EL PAIS COEFICIENTE DE IMPACTO SIGNIFICANCIA ESTADISTICA DEPARTAMENTOS CON PENETRACION >25 COEFICIENTE DE IMPACTO SIGNIFICANCIA ESTADISTICA DEPARTAMENTOS CON PENETRACION <25 COEFICIENTE DE IMPACTO SIGNIFICANCIA ESTADISTICA IMPACTO EN EL PIB O.O37 1% % % IMPACTO EN INGRESO MEDIO REAL DEL HOGAR % % % IMPACTO EN LA TASA DE EMPLEO % % % Fuente: Elaboración de Telecom Advisory Services, LLC 48

49 Al analizar los impactos estimados en el PIB del pais En el contexto de otros estudios, se confirma el efecto de retorno a escala Contribucion al PIB si la variable aumenta de 10 puntos porcentuales BANDA ANCHA Y CRECIMIENTO ECONOMICO y = x Brasil R 2 = America La,na Chile OECD- Baja penetracion Alemania- Baja penetracion OECD- Media penetracion Alemania- Alta penetracion OECD- Alta penetracion Fuente: Elaboracion de Telecom Advisory Services Como se observa en el gráfico, la contribución de la banda ancha al PIB se acrecienta con la penetración Esto indica el efecto de aumento de retornos a la inversión de infraestructura como resultado de los efectos de red A par>r de ello se deduce que existe un incen>vo natural a aumentar la penetración de banda ancha 49

50 A partir de esto, corresponde evaluar cual va a ser el impacto economico resultante del alcance de las metas de banda ancha El Plan propone mul>plicar por 4 veces el número de conexiones a Internet de alta velocidad hacia el año 2014 El obje>vo mencionado implica una tasa de crecimiento anual del 41.42% de conexiones de banda ancha a nivel nacional Esto significa aumentar el numero de accesos en 6,600,000, lo que resulta en un nivel esperado de penetración de 18.46% de la población (del 4.83% actual) y 67.58% de los hogares (de 17.7% actual) 50

51 Considerando que un aumento de 10% en la penetración tiene un impacto en el PIB de p.p., esto impactaria el crecimiento anual en 0.15% El crecimiento anual de la penetración de banda ancha del 41,42% a nivel nacional resultaría en un crecimiento anual del PIB de 0.15% (6% de lo que fue el crecimiento en el año 2010) Esto implica una contribución al PIB entre el 2010 y el 2014 del orden de US $ 2,685 millones IMPACTO EN EL PIB DEL PLAN DE BANDA ANCHA $435,000 $434,500 $434,000 $434,585 $433,500 $433,000 $432,500 $432,000 CAGR: 0.15% $431,900 $431,500 $431,000 $430, Fuente: Elaboracion de Telecom Advisory Services El valor mencionado debe ser considerado como un piso del impacto debido a que no incluye efectos indirectos, como los que pueden darse a través de la mejora en el nivel de empleo (e.g. empleos inducidos) 51

52 El impacto economico del plan puede resumirse asi: $ 2.7 billones de PIB, creacion de 230,000 empleos y aumento del ingreso de 0.15% anual ESTIMACION DE IMPACTO ECONOMICO DE LA DIGITALIZACION VARIABLE IMPACTO Número de Conexiones 2,200,000 8,800,000 6,600,000 Individuos 45,509,584 47,661,787 2,152,203 Penetración Individuos (%) 4.83 % % % Hogares 12,427,272 13,022, ,074 Penetración Hogar (%) % % % Empleos 19,585,836 20,522, ,646 (*) PBI (Millones USD) 431, ,585 2,685 Ingreso real por hogar $ 9, $9, $ (*) Empleos anuales Fuente: Elaboracion de Telecom Advisory Services 52

53 Contents ICT Evidence- based policy making and the need of data Measuring broadband impact: data and analy>cal problems A case study of economic impact measurement of a na>onal broadband plan The next fron>er in ICT impact analysis: digi>za>on 53

54 Digitization is the capacity to use digital technologies to generate, process, share and transact information To achieve a significant impact, digi>za>on has to be widely adopted in the economic and social fabric of a given na>on U>lized by individuals, economic enterprises and socie>es Embedded in processes of delivery of goods and services Relied upon to deliver public services For this condi>on to occur, digi>za>on has to fulfill several condi>ons Affordable: it has to be sufficiently affordable to allow scalable impact Ubiquitous: it is supported by telecommunica>ons networks and reaching most popula>on of a na>onal territory Accessible: networks need to be accessed by mul>ple fixed and mobile voice and data devices Reliable: networks have to provide sufficient capacity to deliver vast amounts of informa>on at speeds that do not hinder their effec>ve use 54

55 A composite index capturing 23 indicators was developed to measure the level of digitization of a given country SUB-INDEX METRIC INDICATORS Affordability Reliability Pricing of different telecommunications services that make digitization accessible to the population and enterprises Level of resilience of networks that transport digital information Average of fixed, mobile and broadband access and usage costs, adjusted for GDP per capita Fixed costs include line installation cost, and fixed line cost per minute; mobile costs include mobile connection fee and mobile prepaid tariffs; broadband costs correspond to the broadband tariffs Measured by the overall mobile, broadband and fixed investments per telecom subscriber Access Adoption of devices and telecommunications networks that allow consumers and enterprises access to digital networks Averages the sum of fixed and mobile broadband, mobile phone and PC penetrations, and the coverage of mobile networks Capacity Network throughput indicating quality of service Measures the international internet bandwidth and the nominal downloading broadband speeds offered to subscribers Usage Adoption of ICT platforms and changes in business processes indicating assimilation of digital technologies Average of E-commerce penetration, E-Government services offered, data usage, online entrepreneurship, social media usage, percentage of individuals using the internet, and data as a percentage of wireless ARPU Human Capital The skill level of the workforce to use ICT Percent of engineers in the total employed population, and the percent of population with tertiary education 55

56 We observe also that, while being highly related to broadband, digitization is measuring something more than broadband adoption Fixed Broadband Penetration DIGITIZATION INDEX VERSUS BROADBAND PENETRATION 50 y = x R 2 = Digitization Index 56

57 The 2010 digitization index was calculated for 157 countries, revealing four clearly development stages 70 STAGES OF DIGITIZATION DEVELOPMENT Norway Korea US Hong Kong Switzerland Taiwan Luxemburg Canada Israel Iceland Japan Sweden UK Finland Australia Belgium 1 9 N. Zealand Denmark France Singapore Portugal Netherlands Germany Austria ADVANCED TRANSITIONAL EMERGENT LIMITED Spain Poland Estonia Oman Peru S. Africa Sri Lanka Italy Slovenia Bulgaria Bahrain Colombia Lebanon Pakistan Russia Hungary Qatar Philippines Belarus Indonesia Thailand Czech R. Lithuania Latvia Iran Venezuela Tunisia Namibia Ireland U.A.E. Croa,a Turkey Botswana Egypt Morocco Slovak R. Chile Uruguay Mexico Ukraine Mongolia Kyrgyzstan Greece S. Arabia Kuwait Costa Rica China India Angola Romania Malaysia Argen,na Brasil Jordan Algeria Syria Ghana C. d Ivoire Swaziland Nicaragua Nigeria Bangladesh Afghanistan Kenya 57

58 Access and affordability are less of a world problem; big spreads between mature and emerging countries are usage, capacity and human capital Norway Luxembourg United France Spain Greece Macao, China Mauritius Croatia Antigua and Mexico Serbia T.F.Y.R. Ecuador Maldives Gabon Mongolia Pakistan Thailand Kyrgyzstan Cambodia Zambia Tanzania Cuba Cameroon Eritrea Ethiopia AFFORDABILITY INFRASTRUCTURE RELIABILITY NETWORK ACCESS CAPACITY USAGE HUMAN CAPITAL TOTAL For all countries, normalized usage sub- index never matches the levels of access sub- indices, which indicate a big challenge across the world For all OECD and middle income countries, the sub- indices affordability and network access tend to be consistently above the digi,za,on index indica,ng that countries have addressed the access challenge The affordability and capacity sub- indices tend to rapidly drop at low GDP levels 58

59 The analysis of countries that are rapidly increasing their digitization could be linked to two distinct economic development drivers EMERGING COUNTRIES: COMPARATIVE EVOLUTION OF DIGITIZATION ( ) The country s economy is based on commerce and services, which require digi,za,on to become more efficient Singapore India Brazil UAE Economic growth came ini,ally from manufacturing, but the spread of this expansion required some level of digi,za,on in order to con,nue growing and to start migra,ng to a more developed economy Egypt Kuwait Qatar China 59

60 An econometric model was built to test the contribution of GDP to economic growth Built a panel of 18 countries and u>lized a produc>on func>on model which includes digi>za>on, human capital and capital forma>on with the purpose of understanding the rela>ve impact of digi>za>on on economic development Countries included in the model: United States, United Kingdom, Japan, Singapore, Korea, Australia, Germany, Norway, Sweden, India, China, U.A.E., Kuwait, Brazil, Indonesia, Qatar, and Egypt Data: Panel with observa>ons ranging between 1995 and 2010 (206 observa>ons) Dependent variable: GDP Independent variables: the digi>za>on index, educa>on (as secondary school enrollment), and gross capital forma>on 60

61 The model states that for developed and developing countries 1 point increase on digitization increases GDP per capita by 0.050%. reg lgdp lindex lcapital leducation1 yr_* cid_* Source SS df MS Number of obs = 206 F( 33, 172) = Model Prob > F = Residual R-squared = Adj R-squared = Total Root MSE = lgdp Coef. Std. Err. t P> t [95% Conf. Interval] lindex lcapital leducation

62 Model results are robust and coefficient of index is statistically significant One point increase on digi>za>on increases GDP per capita growth by percentage points Given the fact that economies are typically growing at a rate ranging between 4% and 5% (or even less), the level of contribu>on of digi>za>on to growth is extremely important Regarding the results, the coefficients for each variable go in the right direc>on (posi>ve in digi>za>on, capital forma>on and human capital) Capital and labor contribu>on reflect appropriate orders of magnitude for produc>on func>ons Level of significance of coefficients: Digi>za>on: 1% Educa>on: 1% Gross Capital Forma>on: 1% Country and year fixed effects are included in regression to control for heterogeneity of specific occurrences 62

63 More importantly, digitization has a higher impact on economic development than only broadband 0.6 DIGITIZATION AND ECONOMIC DEVELOPMENT Contribu,on to GDP growth of 10 point increase in variable POINT INCREASE IN DIGITIZATION YIELDS 0.50% INCREASE IN GDP PER CAPITA 0.0 OECD-High OECD-Med. OECD-Low Germany-High Germany-Low L. America Brazil Chile Malaysia Low income High income Broadband Studies Mobile Studies Full economic impact ICT is achieved through the cumula,ve adop,on of all technologies, in addi,on to the assimila,on and usage in the produc,on and social fabric Achieving broadband penetra,on is only one aspect of required policies; maximiza,on of economic impact can only be achieved through a holis,c set of policies ranging from telecoms to compu,ng to adop,on of internet and ecommerce DIGITIZATION 63

64 A second model was built to determine the impact of digitization on changes in the unemployment rate Built a fixed effects panel data regression, 18 countries studied from 1996 to 2010 Countries included: Australia, Brazil, China, Egypt, Germany, India, Indonesia, Japan, Kuwait, Norway, Qatar, Saudi Arabia, Singapore, South Korea, Sweden, United Arab Emirates, United Kingdom, United States Data u>lized: Dependent variable: unemployment rate between year t and t+1 Control variables: Gross Capital Forma>on, Foreign Direct Investment, percent of service imports that are insurance or finance related, secondary school enrollment Model was run on 172 observa>ons 64

65 This model states that a 10 point increase on digitization increases leads to a -.82% change in unemployment rate. reg unemployment index grosscapitalformationofgdp schoolenrollmentsecondarygross foreigndirectinvestmentnetinflow insura > nceandfinancialservicesofc yr_* countryid Source SS df MS Number of obs = 172 F( 20, 151) = 4.17 Model Prob > F = Residual R-squared = Adj R-squared = Total Root MSE = unemployment Coef. Std. Err. t P> t [95% Conf. Interval] index grosscapit~p schoolenro~s foreigndir~w insurancea~c

66 The econometric impact model on job creation is also robust All control variables have the right sign All canonical variables are sta>s>cally significant The model is very stable, meaning that signs and sta>s>cal significance of each of the coefficients are stable even when the specifica>ons are changed By including GDP per capita in the variables, we negate the possibility that the index is working as a proxy for level of development 66

67 As in the case of GDP growth, digitization has a higher impact on job creation than broadband Contribu,on to job crea,on of 10 point increase in variable US - Metro Areas 0.30 Germany-High 0.06 DIGITIZATION AND EMPLOYMENT Germany-Low 0.03 Mexico 0.38 Argentina Broadband Studies 0.29 Brazil 0.45 Chile 0.02 DIGITIZATION POINT INCREASE IN DIGITIZATION YIELDS 0.82% DECREASE IN UNEMPLOYMENT RATE Full deployment and assimila,on of ICT has a much larger impact on employment because it contributes to more jobs in the ICT sector (sogware development, Business Process Outsourcing, equipment manufacturing and parts supplies) In addi,on, the impact of assimila,on of ICT through enhanced usage has spill- over impact on other sectors of the economy (in par,cular, trade, financial services, health care) 67

68 68

Sitemap