Impact of tax burden on regional productivity growth

Transcription

1 Id539 / 1 Impact of tax burden on regional productivity growth OECD Workshop: Effective Corporate Taxation Martin Eichler, BAK Basel Economics July 4 th, 2006, OECD, Paris

2 Id539 / 2 How relevant is policy for regional growth? Policy question: What are the drivers of regional growth? What are the measures decision makers can choose to foster economic growth? BAK Analytical Approach: 1. Benchmarking: Comparing economic performance and location factors 2. Analysis: Understanding the relationship between performance and location factors 3. Policy: Advice on policy measures and possible trade-offs BAK research project: Policy and Regional Growth Understanding the relation between location factors and regional growth using empirical and econometric tools Project Sponsors:

3 Growth of productivity per man hour Selected European countries % 4% 3% 2% 1% 0% DE FK IT AT CH UK Id539 /

4 Id539 / 4 Growth of productivity per man hour Variation of German regions, % DE 4% DE_MIN DE_MAX 3% FK 2% IT AT 1% CH UK 0%

5 Id539 / 5 Growth of productivity per man hour Variation of UK regions, % DE 4% DE_MIN DE_MAX FK 3% IT AT 2% CH UK 1% UK_M IN UK_M AX 0%

6 Id539 / 6 Growth of productivity per man hour Variation in selected European countries, % DE DE_MIN DE_MAX 4% FK FK_MIN FK_MAX 3% IT IT_M IN IT_M AX AT 2% AT_MIN AT_MAX CH CH_MIN 1% CH_MAX UK UK_M IN UK_M AX 0%

7 Id539 / 7 Database International Benchmarking Database Economic Performance ( ) Indicators: GDP Gross value added Employment Hours worked Productivity (hourly productivity and output per employee) Measures: Gross and real Level and growth (time series 1980 to 2005) In national currency and in common currency Total, per capita Industries: 60 different industries (following NACE 2-digit level)

8 Id539 / 8 Database International Benchmarking Database Location Factors Innovation: Expenditure on R&D* ( ) University quality (Shanghai Index) ( ) Number of higher education degrees* (tertiary degrees) ( ) Education level of labour force* (secondary / tertiary) ( ) Patents* ( ) Bibliometric indicators* ( ) Company start-ups* ( ) Taxation: Taxation of highly qualified employees ( ) Taxation of companies ( ) Regulation: Regulation of product markets ( ) Regulation of labour markets ( ) * Indicators by industries, directly or through subjects/technology fields. Some parts of database still in development.

9 Id539 / 9 Database International Benchmarking Database Location Factors Accessibility: Global accessibility ( ) Regional accessibility ( ) Labour costs: Wages costs ( ) Unit labour costs ( ) Population: According to gender and different age cohorts ( ) Agglomeration: Agglomeration size Proximity to an agglomeration

10 Id539 / 10 International Benchmarking Database 2006 Regional coverage

11 Id539 / 11 Database Estimation data Data selection Left hand variable: Productivity (GDP, employment) growth, annual (alternative: longer periods) Transformations: Lagged Moving averages Derivations from sample average Levels or changes of location factors Controls: Country dummies (country groups) Time dummies Global and Structural (industry mix) Effect Economic cycle

12 Id539 / 12 GDP growth and taxation manpower: evidence Distribution of GDP growth and tax burdens on income of highly qualified employees 6% 5% 4% 3% 2% 1% 0% 25% 30% 35% 40% 45% 50% 55% 60% Source: BAK International Benchmarking Database 2006 GDP growth ( annual average) Taxation: income , single, level in 1995

13 Id539 / 13 GDP growth and taxation companies: evidence Distribution of GDP growth and tax burdens on company investments 6% 5% 4% 3% 2% 1% 0% 10% 15% 20% 25% 30% 35% 40% 45% 50% 55% Source: BAK International Benchmarking Database 2006 GDP growth ( annual average) Taxation: company, standard investment, level in 1995

14 Id539 / 14 Why does a region grow? Econometric analysis Theory and observation (questionnaires) can tell which location factors are important but they do hardly tell the size of the impact on regional growth Econometric analysis based on the benchmarking database - Basic equation: Growth it = α 0 + α 1 *Global Effect t + α 2 *Structural Effect it + β 1 *L 1it-1 + β 2 *L 2it ε it - Panel data estimation (random effects, fixed effects) with i = 142 regions t = 15 years ( ) several location factors (L)

15 Id539 / 15 Econometric results Coefficients (annual productivity growth, ) Variable Variable definition coefficient prob. (β i =0) Global Effect overall sample productivity growth + < 1% Structural Effect productivity growth for given industry mix + < 1% Company Taxation 4 years moving average, relative to sample mean < 5% Manpower Taxation 4 years moving average, relative to sample mean < 1% Labour Market Regulation 4 years moving average, relative to sample mean + < 1% Product Market Regulation 4 years moving average, relative to sample mean + < 1% Innovation: Secondary Educ. difference, 4 years moving average + < 1% Innovation: Tertiary Educ. difference, 4 years moving average + < 10% Innovation: R&D Expen. difference, 4 years moving average + < 5% Accessibility: Intercontinental difference, 10 years moving average + < 10% Accessibility: Interregional difference, 10 years moving average < 1% Regional Spill Over time invariant < 5%

16 Id539 / 16 Impact of policy variables: taxation Estimated impact of taxation on productivity growth in selected regions 6% 4% Taxation of companies Taxation of highly qualified employees 2% 0% -2% -4% -6% -8% Bruxelles København München Frankfurt Madrid Paris Dublin Milano Wien Basel Zürich Helsinki Stockholm London Glasgow Source: BAK International Benchmarking Database 2005 Impact on productivity growth, in %-points, cumulated

17 Id539 / 17 Impact of location factors Sample-distribution (regions), impact on productivity growth ( ) Intercontinental Accessibility Innovation - R&D expenditures Innovation - tertiary education Innovation - secondary education Taxation of Companies Spatial Spillover Effect Interregional Accessibility Regulation of Labour Markets Taxation of highly qualified employees Structural Effect 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% Coefficient negative Source: BAK International Benchmarking Database 2005 Impact on productivity growth, cumulated (in %-points, max impact against min impact)

18 Id539 / 18 Research agenda 2006 Regional coverage US states Increase regional coverage in Europe (France, Italy), Portugal and Greece Left hand side Include labour market (hours worked, employment ratio, population) Industry specific estimations Methods Evaluation of system approach Right hand side

19 Id539 / 19 Impact of location factors Estimated impact of taxation on GDP growth in selected regions 1.5% 1.0% taxation taxation of highly qualified employees taxation of companies 0.5% 0.0% -0.5% -1.0% Zurich Dublin London Oslo Vienna Boston San Francisco Helsinki Madrid Paris Brussels Copenhagen Amsterdam Stockholm Munich Milan Stuttgart Source: BAK International Benchmarking Database 2006 Impact on GDP growth, in %-points, annual average

20 Id539 / 20 Impact of location factors Sample-distribution (regions), the annual impact on real GDP growth percentiles 0-10 % % % % % innovation r&d innovation human capital taxation companies taxation highly-qualified employees regulation labour markets regulation product markets structural effect -1.5% -1.0% -0.5% Median 0.0% 0.5% 1.0% 1.5% 2.0% Source: BAK International Benchmarking Database 2006 Impact on GDP growth, annual average (in %-points, derivation from sample mean)

21 Id539 / 21 Summary of results OECD workshop on corporate taxation, July 4 th, 2006 Research program on impact of location factors on economic growth Regional variation is large and available location factors can only explain a part of the difference Signs of location factors largely as expected from theory & significant Strong country effects and time trend, industry mix has important influence Impact of taxation Tax burdens exercise ceteris paribus a negative impact on economic growth This result is confirmed by stability tests Differences of the tax burden on the income of highly qualified employees explains more regional growth differences than company taxation, especially for productivity growth

22 Id539 / 22 Research agenda OECD workshop on corporate taxation, July 4 th, 2006 Further research directions Regional coverage More specific on sources of growth (productivity, hours worked, employment ratio, population) Advanced Methods (system estimation) Add more location factors, especially varying between regions (regulation; fragmentation and decentralization; innovation; ) Industry specific estimations Further research directions (taxation) Test more different types of tax burdens Influence on specific industries different? Add more regions with different tax burdens