COINVEST Competitiveness, Innovation and Intangible Investment in Europe Intangible investments in Portugal The value of Training Francisco Lima, IST Lisbon, March 18-19, 2010 Project funded by the European Commission under the Seventh Framework Programme Grant No

Plan 1. Intangible investments in Portugal 2. The value of training

1. Intangible investments in Portugal Francisco Lima IST and CEG-IST, Technical University of Lisbon Miguel Torres Preto IST and IN+, Centre for Innovation, Technology and Policy Research, Technical University of Lisbon Pedro Faria University of Groningen, The Netherlands and IN+, Centre for Innovation, Technology and Policy Research, IST

Data sources EU KLEMS Matched employer-employee data (QP – Quadros de Pessoal) National Accounts Community Innovation Survey (CIS)

Computerized Information Computer software – EUKLEMS database Includes own use, purchased and custom software Computerized databases – QP - Database Management sector (NACE K72400)

Innovative property Scientific R&D (includes R&D in social science and humanities) – BERD (OECD - Main Science and Technology Indicators ) Mineral exploration – QP - Mineral exploration industry (NACE excluding “Services associated to oil and gas extraction not related to prospection” and “Salt refinement”) Copyright and license costs – QP - TV and radio, publishing, music industries, acting activities, reproduction of recorded sounds or videos New product development costs in the financial industry – QP - Turnover from the financial industry * share of innovation expenditures of the financial industry obtained from the CIS III (4%) (update with CIS IV) New architectural and engineering design – QP – 50% of the turnover from the architectural and engineering industry plus the earnings from designers in other industries

Economic competencies Advertising expenditure – QP - Advertising industry (NACE 744) Market research – QP - Market research industry Firm-specific human capital – Compensation of employees (EU Klems) * Percentage of training costs obtained from the Eurostat Vocational Training Survey (1999 value) Organizational structure - Purchased – Quadros de Pessoal - Management consulting industry Organizational structure Own Account – QP – 20% of managers' earnings

Intangible investment (Million EUR)

Intangible Investment in the Market Sector (% GDP)

Growth Accounting without Intangible Assets Annual growth rate of labor productivity of the business sector 2.76%0.91% Contribution of Inputs ICT tangible capital deepening0.57%0.51% Non-ICT tangible capital deepening1.26%0.88% Labor quality-0.22%0.83% MFP1.07%-1.38% Growth Accounting with Intangible Assets Annual growth rate of labor productivity of the business sector3.06%1.20% Contribution of Inputs ICT tangible capital deepening0.45%0.44% Non-ICT tangible capital deepening0.73%0.71% Intangible capital deepening-0.38%0.64% Labor quality-0.20%0.74% MFP2.32%-1.40% Software-0.42%0.07% Innovative property0.16%0.32% Economic competence-0.11%0.25%

What’s next? Improve the estimates Extend the period of analysis Other sources/methodologies Matching CIS with firm data / CIS with matched employer- employee data ECVTS IPCTN - Survey on the National Science and Technology Potential More on occupations – matched employer-employee data

2. The value of training Francisco Lima Susana Neves, INE – Statistics Portugal

Training Intermediate expenditure or investment? Affects productivity? Reflected in wages? Objective: use the recent Adult Education Survey (AES) to measure the relationship between training and wages – adult participation in formal and non-formal education and training and informal activities

Adult Education Survey INE - Statistics Portugal, 2007 – conducted in all European Union Member States, following Eurostat’s methodological guidelines) observations, 5741 employees Wages coded – use interval regression individuals aged between 18 and 64 years old – 23.1% were involved in non formal education and training activities

Wage equations Log wage – estimated using interval regression Model 1 – Training, education, tenure, age, region Model 2 – Variables from Model 1 plus occupation, industry and firm size

Problems? Education and training endogenous? (Omitted) ability or human capital? Measurement error (less likely, at least for education) Use instrumental variables

Instruments Individuals were asked about parents’ education and labor market status – employed, unemployed, inactive – when aged In addition, there is also information about – Informal learning activities – Reading habits - newspapers – Books at home – Participation in several type of organization activities – religious, political, cultural – Attendance to cultural and sports events

Model 1

Model 2

Discussion Given the expected upward bias (missing “ability”), unexpected increase in the estimated coefficients Or correcting an attenuation bias as a result of measurement error? Weak instruments Existence of heterogeneity in individual returns – Instruments that influence only the educational decision of individuals with high marginal returns due to either liquidity constraints or to high ability (Card, 1995) Local average treatment effect (LATE) interpretation – IV identifies only the average returns of those who comply with the assignment-to-treatment mechanism implied by the instrument (Imbens and Angrist, 1994)

Conclusion Training (and education) – associated with higher wages The benefits of investing in training spread for several periods (higher marginal product) – otherwise, no relationship with wages – given the estimates for a sample of the population of workers Investment More work – able to improve the estimates with the information at hand?