El documento describe un estudio que analiza los diferentes estadios de desarrollo digital de países y los factores que determinan su pertenencia a cada estadio. El estudio identifica cuatro estadios (líderes digitales, seguidores digitales, rezagados digitales y saltadores digitales) y analiza las características e indicadores clave de cada uno. Además, realiza regresiones logísticas para determinar los principales predictores de pertenencia a los estadios de líderes digitales y rezagados digitales.
Midiendo el Desarrollo Digital para las Políticas Públicas:el Papel del Gobierno
1. Midiendo el Desarrollo Digital para las Políticas Públicas: el Papel del Gobierno Ismael Peña - López Universitat Oberta de Catalunya II Conferencia Internacional sobre Brecha Digital e Inclusión Social Leganés, 29 de Octubre de 2009
2. Hipótesis Estadios de desarrollo digital Riqueza y desarrollo económico Educación Infraestruc-turas digitales /43 Compromiso del Gobierno con las TIC Régimen de Incentivo de la Economía
3.
4.
5. Modelo: 360º Digital Framework /43 Fuente: autor Activos Flujos Oferta Demanda Infrastructuras Disponibilidad Asequibilidad Sector TIC Empresas Economía Recursos Humanos Marco Legal Regulación (Sector) TIC Estrategias y Políticas Sociedad Información Contenido y Servicios Disponibilidad Intensidad de Uso Competencias Digitales Nivel de Alfabetitzación Digital Alfabetización Digital (Formación)
6. De la teoría a la práctica Indicadores (vars.) utilizadas para la caracterización de los estadios de desarrollo digital (WITSA) Indicadores (luego variables) utilizadas para los conglomerados(WITSA) /43 Infrastruct. Sector TIC Alfabetiz. Digital Marco político y regulatorio Contenidos y Servicios No digital Oferta/Activos 8 2 2 3 5 27 Demanda/Flujos 5 4 1 2 6 Infrastructuras Sector TIC Alfabetización Digital Marco político y regulatorio Contenidos y Servicios Oferta/Activos 6 1 1 2 3 Demanda/Flujos 1 1 1 1 5
7. Centros de conglomerados (WITSA) 1 - Broadband subscribers (per 100 people) 2 - Personal computers (per 100 people) 3 - Telephone mainlines (per 100 people) 4 - Mobile phone subscribers (per 100 people) 5 - International Internet bandwidth (bits per person) 6 - Internet Hosts (per 10000 people) 7 - Price basket for residential fixed line (US$ per month) 8 - Telecommunications revenue (% GDP) 9 - GDP per Telecom Employee (US Dollars) 10 - Human Capital 11 - Internet Access in Schools 12 - Laws relating to ICT 13 - Intellectual property protection 14 - Gov't procurement of advanced tech products 15 - Secure Internet servers (per 1 million people) 16 - Total Domains (per 100 people) 17 - Availability of government online services 18 - Internet users (per 100 people) 19 - Total ICT Spending, Consumer (% of GDP) 20 - Firm-level technology absorption 21 - Extent of business Internet use 22 - ICT use and government efficiency Análisis de conglomerados no jerárquicos de K-medias. Significatividad de la F en la ANOVA para todas las variables: p<0.001 /43
8.
9. Infraestructuras 1 - Broadband subscribers (per 100 people) (*) 2 - Personal computers (per 100 people) (*) 3 - Telephone mainlines (per 100 people) (*) 4 - Mobile phone subscribers (per 100 people) (*) 5 - Population covered by mobile telephony (%) (*) 6 - International Internet bandwidth (bits per person) (*) 7 - Internet Hosts (per 10000 people) (*) 8 - Internet subscribers (per 100 inhabitants) (*) 9 - Residential monthly telephone subscription (US$) (**) 10 - Price basket for Internet (US$ per month) (**) 11 - Price basket for mobile (US$ per month) (**) 12 - Price basket for residential fixed line (US$ per month) (*) 13 - Telephone average cost of call to US (US$ per three minutes) (***) % de países que puntuaron “alto” en el indicador por conglomerado (*): p<0.01 (**): p<0.05 (***): p<0.1 /43 Leaders Laggards
10. Sector TIC 1 - Telecommunications revenue (% GDP) (*) 2 - High-technology exports (% of manufactured exports) (**) 3 - Telephone subscribers per employee (***) 4 - Telephone employees (per 100 people) (**) 5 - Total full-time telecommunications staff (per 100 people) (*) 6 - GDP per Telecom Employee (US Dollars) (*) % de países que puntuaron “alto” en el indicador por conglomerado (*): p<0.01 (**): p<0.05 (***): p<0.1 /43 Leaders Laggards
11. Alfabetización Digital 1 - Enrolment in science. Tertiary. (per 100 people) (*) 2 - Human Capital (*) 3 - Internet Access in Schools (*) % de países que puntuaron “alto” en el indicador por conglomerado (*): p<0.01 (**): p<0.05 (***): p<0.1 /43 Leaders Laggards
12. Marco político y regulatorio 1 - Laws relating to ICT (*) 2 - Intellectual property protection (*) 3 - Level of competition - DSL (**) 4 - Level of competition – Cable modem (**) 5 - Gov't procurement of advanced tech products (*) % de países que puntuaron “alto” en el indicador por conglomerado (*): p<0.01 (**): p<0.05 (***): p<0.1 /43 Leaders Laggards
13. Uso 1 - Secure Internet servers (per 1 million people) (*) 2 - Total Domains (per 100 people) (*) 3 - Total ICT Spending, Retail Trade (% of GDP) (*) 4 - Web Measure (*) 5 - Availability of government online services (*) 6 - International outgoing telephone traffic (minutes) (per 100 people) (*) 7 - Internet users (per 100 people) (*) 8 - E-Participation (*) 9 - Total ICT Spending, Consumer (% of GDP) (*) 10 - Firm-level technology absorption (*) 11 - Extent of business Internet use (*) % de países que puntuaron “alto” en el indicador por conglomerado (*): p<0.01 (**): p<0.05 (***): p<0.1 /43 Leaders Laggards
14. Indic. Analógicos 1 - GDP (***) 2 - GDP Capita (*) 3 - GDP per capita, PPP (current international $) (*) 4 - GNI per capita, Atlas method (current US$) (*) 5 - GNI per capita, PPP (current international $) (**) 6 - HDI (*) 7 - Life expectancy at birth, total (years) (*) 8 - Improved water source (% of population with access) (*) 9 - Health Public Expenditure (% of govt. expenditure) (*) 10 - Health Public Expenditure (% of total Health expend.) (*) 11 - School enrollment, primary (% net) (***) 12 - School enrollment, primary (% gross) (**) 13 - Education Public Expenditure (% of govt. expenditure) (***) 14 - Gross National Expenditure (% of GDP) (**) 15 - General Govt. final consumption expend. (% of GDP) (***) 16 - Economic Incentive Regime (*) 17 - Innovation (*) 18 - Population in urban agglom. > 1 million (% of total pop.) (*) 19 - Inequality-10 (**) 20 - Mortality rate, infant (per 1,000 live births) (*) 21 - Population growth (annual %) (***) 22 - Interest payments (% of GDP) (*) 23 - Present value of debt (% of GNI) (**) 24 - GDP deflator (base year varies by country) (*) 25 - Inflation, consumer prices (annual %) (*) 26 - Inflation, GDP deflator (annual %) (*) 27 - Tax revenue (% of GDP) (**) % de países que puntuaron “alto” en el indicador por conglomerado (*): p<0.01 (**): p<0.05 (***): p<0.1 /43 Leaders Laggards
15. Determinantes: líderes digitales /43 logit(ZCLUSTER54_CB) = β1 • GEN30 + β2 • GEN05 + β3 • GEN07 + β4 • GEN08 + β5 • LEGAL_D_04+ ε Regresión logística binaria para los digital leaders (1 es un digital leader, 0 no es un digital leader) como variable dependiente . B S.E. Wald df Sig. Exp(B) Life expectancy at birth, total (GEN30) -.399 .208 3.664 1 .056 .671 Inequality-20 (GEN05) -1.066 .578 3.403 1 .065 .344 Urban Population (%) (GEN07) .138 .079 3.030 1 .082 1.148 Economic Incentive Regime (GEN08) 1.671 .877 3.628 1 .057 5.317 Government prioritization of ICT (LEGAL_D_04) 2.869 1.737 2.727 1 .099 17.611 N 46 Correctly predicted cases 95.7% 96.8% (leaders) 93.3% (resto) -2 Log likelihood 15.970 Cox & Snell R-square .646 Nagelkerke R-square .862 Chi-Square (sig) 47.799 (.000) Hosmer and Lemeshow Test Chi-Square (sig) 1.546 (.981)
16. Determinantes: rezagados digitales /43 logit(ZCLUSTER54_CBL) = β0 + β1 • GEN06 + β2 • GEN14 + β3 • INF_S_06 + β4 • LEGAL_D_01 + ε Regresión logística binaria para los digital laggards (1 es un digital laggard, 0 no es un digital laggard) como variable dependiente . B S.E. Wald df Sig. Exp(B) Constant 38.214 16.958 5.078 1 .024 3.945·10 16 Inequality-10 (GEN06) -.235 .138 2.909 1 .088 .790 Health Public Expenditure (% of total Health expenditure) (GEN14) -.176 .081 4.665 1 .031 .839 Population covered by mobile telephony (%) (INF_S_06) -.100 .050 3.936 1 .047 .905 Importance of ICT to government vision of the future (LEGAL_D_01) -4.304 2.239 3.696 1 .055 .014 N 47 Correctly predicted cases 94.6% 96.4% (laggards) 88.9 % (resto) -2 Log likelihood 11.391 Cox & Snell R-square .551 Nagelkerke R-square .823 Chi-Square (sig) 29.663 (.000) Hosmer and Lemeshow Test Chi-Square (sig) 3.684 (.815)
17. Conclusiones Estadios de desarrollo digital Riqueza y desarrollo económico Educación Infraestruc-turas digitales Estadios de desarrollo digital Riqueza y desarrollo económico Educación Infraestruc-turas digitales Compromiso del Gobierno con las TIC Régimen de Incentivo de la Economía Compromiso del Gobierno con las TIC Régimen de Incentivo de la Economía
18. Conclusiones Estadios de desarrollo digital Riqueza y desarrollo económico Educación Infraestruc-turas digitales Estadios de desarrollo digital Riqueza y desarrollo económico Educación Infraestruc-turas digitales Compromiso del Gobierno con las TIC Régimen de Incentivo de la Economía Compromiso del Gobierno con las TIC Régimen de Incentivo de la Economía Estadios de desarrollo digital Igualdad económica Educación Salud Priorización de las TIC en el gob. Importancia de las TIC en la visión del futuro del gobierno Leapfroggers
19.
Notas del editor
WH3: Higher levels of wealth and economic development, education and the existence of digital infrastructures almost always coincide with higher levels of digital development. Nevertheless, Governments can accelerate the process of digital development through the adoption of public policies that frame and foster the Information Society – such as Government prioritization of ICT and assigning a high importance to ICT in government vision of the future – and establishing an appropriate Economic Incentive Regime. This will raise the probability of a country of reaching higher stages of digital development.
Analysis of variables to avoid problems of multicollinearity Year 2007 really means that they were collected during 2007, but published afterwards. Sometimes data come from previous years. On the other hand, 2007 is the last year of an economic cycle: taking 2008 would surely change things. It is also the year of social networking sites. Problem of not having time series. Aggregation per country blurs the details within countries. Ways to simplify information: Factor analysis (non conclusive) and cluster analysis We repeated some statistics using two different datasets: WITSA and OECD. The former gathers the 75 most digitally developed countries which, in some way, are all the digitally developed countries of the World, leaving outside almost 200 countries that are simply too far from being called digital. The later, the OECD data set, includes arguably the most developed countries in the World. By using this second dataset we want to zoom into wealth and see whether, at the macro aggregate level, the patterns found for the WITSA dataset (broad range of economies) still apply when finding differences within a narrower and richer range of economies.
Analysis of variables to avoid problems of multicollinearity Year 2007 really means that they were collected during 2007, but published afterwards. Sometimes data come from previous years. On the other hand, 2007 is the last year of an economic cycle: taking 2008 would surely change things. It is also the year of social networking sites. Problem of not having time series. Aggregation per country blurs the details within countries. Ways to simplify information: Factor analysis (non conclusive) and cluster analysis We repeated some statistics using two different datasets: WITSA and OECD. The former gathers the 75 most digitally developed countries which, in some way, are all the digitally developed countries of the World, leaving outside almost 200 countries that are simply too far from being called digital. The later, the OECD data set, includes arguably the most developed countries in the World. By using this second dataset we want to zoom into wealth and see whether, at the macro aggregate level, the patterns found for the WITSA dataset (broad range of economies) still apply when finding differences within a narrower and richer range of economies.
Why not analogue indicators Infrastructures: Information and Communication Technologies. Can be divided into three groups: hardware, software and connectivity. Infrastructures, Availability: the presumed existence of these infrastructures. Infrastructures, Affordability: the cost for the end user of the acquisition of such infrastructures in relationship with one individual or community’s economic power – hence, the price in real terms. ICT Sector: The economic sector responsible for the provision of ICT Infrastructures ICT Sector, Enterprises / Industry: The existence of firms whose activities fits within the definition of the ICT sector. ICT Sector, Workforce: Skilled employees that work directly in the ICT Sector or whose activities are closely related to it . Digital Skills: Skills related relevant both to the use of electronic devices and the use of information in digital format Digital Skills, Digital Literacy Level: The measured levels of such skills in an individual or a community, both in relation to the number of literate people and the degree of their literacy. Digital Skills, Digital Literacy Training: The existence of courses, curricula or other training plans to increase the Digital Literacy Level. Policy and Regulatory Framework: Whether there are explicit rules, laws, policies, etc. that directly affect and try to put in order the Digital Economy. Policy and Regulatory Framework, ICT (Sector) Regulation: Rules created by the Legislative branch or other regulatory bodies to regulate the Digital Economy, especially the ICT Sector and its activities. Policy and Regulatory Framework, Information Society Strategies and Policies: Policies, strategic plans, etc. created by the Executive branch or other functions of government to frame their Digital Economy related policies. Content and Services: Content and services in digital form. Content and Services, Availability: The existence of such contents and services, including those arising from the private sector (for or without profit) and the public sector. Content and Services, Intensity of Use: The use of such content, measured both quantitatively and qualitatively.
How are these tables built – what do they stand for
How is this graphic built Non-hierarchical K-means cluster analysis to segment groups of countries whose variables have statistically significant similar values in opposition to the other groups.
How is this graphic built Contingency tables to find means that are significantly different between clusters and significantly similar within clusters Hypothesis of independence between a chosen variable and the distribution amongst clusters of the country set. A significant score for Pearson Chi-Square and Fischer’s Exact test will reject the hypothesis of independence, meaning that a country’s allocation to a particular cluster depends on its value for that selected variable We measure the correlation of the distribution amongst clusters and that same selected variable by means of the Pearson and Spearman correlations. Again, significant results tell us that both variables (the cluster and the chosen one) are correlated Haberman typified adjusted residuals. These residuals have a normal distribution. Taking a confidence level of 0.95, we can look for adjusted residuals with absolute value over 1.96, noting that there are more (or less ) cases than expected in comparison with the case where the two compared variables (the cluster and the other variable in our case) were independent.
None of the independent variables of the regression were used to build the clusters Why use a probabilistic model, binary, pros and cons Why not strivers and/or leapfroggers The Chi-Square test confirms that the power of the effect of the independent variables taken jointly is statistically significant, and the Hosmer and Lemenshow test rejects the null hypothesis that there is no difference between the observed and predicted values of the dependent variable, thus confirming the goodness to fit of the overall model. Indeed, the model predicts a total of 95.7% of all cases (46 countries), 96.8% of digital leaders and 93.3% of the rest of countries. The high value of Nagelkerke’s R-square implies quite a good degree in the explanatory power of the model too. Life expectancy at birth, total (years) (GEN30): the number of years a newborn infant would live if prevailing patterns of mortality at the time of its birth were to stay the same throughout its life (World Bank, World Development Indicators). Inequality-20 (GEN05): ratio of the richest 20% to the poorest 20% (UNDP, Human Development Report ). Urban Population (%) (GEN07): urban population is the midyear population of areas defined as urban in each country and reported to the United Nations. This indicator measures the proportion between urban and the total population in percent (World Bank, World Development Indicators). Economic Incentive Regime (GEN08): The Economic Incentive and Institutional Regime is the simple average of the normalized scores on three key variables: Tariff & Nontariff Barriers, Regulatory Quality, and Rule of Law (World Bank, Knowledge Assessment Methodology). Tariff & Nontariff Barriers : is a score assigned to each country based on the analysis of its tariff and non-tariff barriers to trade, such as import bans and quotas as well as strict labeling (sic) and licensing requirements (the score is based on the Heritage Foundation's Trade Freedom score and used the World Bank, Knowledge Assessment Methodology) Regulatory Quality : measures the incidence of market-unfriendly policies such as price controls or inadequate bank supervision, as well as perceptions of the burdens imposed by excessive regulation in areas such as foreign trade and business development (World Bank, Governance Indicators / Knowledge Assessment Methodology). Rule of Law : this indicator includes several indicators which measure the extent to which agents have confidence in and abide by the rules of society. These include perceptions of the incidence of both violent and non-violent crime, the effectiveness and predictability of the judiciary, and the enforceability of contracts (World Bank, Governance Indicators / Knowledge Assessment Methodology). Government prioritization of ICT (LEGAL_D_04): measures from 1 (strongly disagree) to 7 (strongly agree) the answer to the question “Information and communication technologies (computers Internet etc.) are an overall priority for the government” (World Economic Forum, Executive Opinion Survey / Global Information Technology Report).
None of the independent variables of the regression were used to build the clusters The Chi-Square test confirms that the power of the effect of the independent variables taken jointly is statistically significant, and the Hosmer and Lemenshow test rejects the null hypothesis that there is no difference between the observed and predicted values of the dependent variable, thus confirming the goodness to fit of the overall model. Indeed, the model predicts a total of 94.6% of all cases (47 countries) – slightly less than the digital leaders model –, 96.4% of digital laggards and 88.9% of the rest of countries. The high value of Nagelkerke’s R-square implies quite a good degree in the explanation power of the model too. Inequality-10 (GEN06): is the ratio of richest 10% to poorest 10% (UNDP, Human Development Report). Health Public Expenditure (% of total Health expenditure) (GEN14): Public Health Expenditure is recurrent and capital spending in Health from central and local governments, external borrowing and grants (including donations from international agencies and nongovernmental organizations), and social (or compulsory) health insurance funds, here measured as percent of total Health Expenditure, which is the sum of public and private health expenditure and covers the provision of health services (preventive and curative), family planning and nutrition activities, and emergency aid for health but excludes provision of water and sanitation. (World Bank, World Development Indicators). Population covered by mobile telephony (%) (INF_S_06): is the percentage of people within range of a mobile cellular signal regardless of whether they are subscribers. (World Bank, World Development Indicators). Importance of ICT to the government vision of the future (LEGAL_D_01): measures from 1 (strongly disagree) to 7 (strongly agree) the answer to the question “The government has a clear implementation plan for utilizing information and communication technologies for improving the country's overall competitiveness” (World Economic Forum, Executive Opinion Survey / Global Information Technology Report).
WH3: Higher levels of wealth and economic development, education and the existence of digital infrastructures almost always coincide with higher levels of digital development. Nevertheless, Governments can accelerate the process of digital development through the adoption of public policies that frame and foster the Information Society – such as Government prioritization of ICT and assigning a high importance to ICT in government vision of the future – and establishing an appropriate Economic Incentive Regime. This will raise the probability of a country of reaching higher stages of digital development.
WH3: Higher levels of wealth and economic development, education and the existence of digital infrastructures almost always coincide with higher levels of digital development. Nevertheless, Governments can accelerate the process of digital development through the adoption of public policies that frame and foster the Information Society – such as Government prioritization of ICT and assigning a high importance to ICT in government vision of the future – and establishing an appropriate Economic Incentive Regime. This will raise the probability of a country of reaching higher stages of digital development.