Andre emner

CS fremkommer i en artikel i dette nummer af Samfundsøkonomen med en lang række kritikpunkter af Velfærdskommissionens analyser og forslag. I det følgende vil vi kommentere de vigtigste af disse påstande og henvise til de relevante sider i Velfærdskommissionens endelige rapport, Velfærdskommissionen (2006).

Jesper Jespersen (JJ) kritiserer i artiklen 'Velfærdskommissionen på et neoklas-sisk vildspor' forskellige forhold, som ikke drejer sig specielt om Velfærdskom-missionens fremskrivninger, men er helt generelle for den makroøkonomiske model DREAM (Danish Rational Economic Agents Model). Der er forskellige fak-tuelle misforståelser i JJs artikel, men derudover er et kernepunkt i JJs kritik, at DREAMs opbygning, ikke mindst af arbejdsmarkedet, overdriver finansieringsbe-hovet som følge af den demografiske udvikling. Hovedpointen hos JJ er, at et større efterspørgselspres fra den offentlige sektors side i sig selv skulle kunne fremskaffe et permanent større beskæftigelsesniveau.

This article presents the results of an analysis of the economic consequences for Denmark of the enlargement of the European Union. Within an applied general equilibrium framework, the effects of the enlargement are quantified with respect to macro economy, production structure and welfare. The relative importance of several elements of the enlargement is illustrated. Specifically, we address the merits of an enlarged single market, repercussions to Danish receipts of transfers from EU funds, enhanced growth in the accession countries, and finally increased immigration of labour. The fully specified enlargement (excluding increased immigration) is demonstrated to give rise to a small long run increase in the domestic level of activity of 0.8 percent and moderate welfare gains amounting to 1.27 percent of 1998 GDP.

The welfare gains from VAT, income tax and corporation tax reductions are evaluated in the dynamic CGE model DREAM. The welfare gains are defined as the sum of the discounted equivalent variations for all current and future generations divided by the discounted value of the lump sum taxes necessary to finance the reduction. We find that a decrease of 5 per cent in each of the tax rates generate welfare gains in the area of 9.49 to 22.94 cents per euro of revenue loss. The different values may indicate a potential welfare gain from a change in the tax structure. However, the intergenerational distribution of the welfare gains differs significantly from one tax experiment to another. This indicates that a change in the tax structure might also generate a shift of welfare between generations. The reduction of VAT and corporation tax increases welfare but involves a generational redistribution of from future generations to current generations. In the case of an income tax cut, welfare increases, and almost all living and future households are better off.

Two macroeconomic models for Denmark; the macroeconometric model ADAM (Annual Danish Aggregate Model) and the dynamic CGE-model DREAM (Danish Rational Economic Agents Model) are compared. An analytical model that contains as special cases a representation of the stationary state in each of the two large simulation models is set up. Analytical multipliers are compared for three different types of policy. In simulations the entire dynamic path of the marginal effects of the same policy changes are compared. Our conclusions are: 1) The qualitative long run effects are similar in the two models. 2) The quantitative effects are in most cases comparable in the long run. 3) The dynamic evolutions of the multipliers are surprisingly similar. A major contributor to quantitative differences is the dissimilarity of the wage curves in the two models: A negative supply effect is the outcome of an increase in the tax burden in DREAM whereas no supply effect is present in ADAM. Second, the price-wage mechanism in ADAM is much stronger than in DREAM. Even if the modelling of private consumption differs significantly between the models, the evolutions of total private consumption in the two models are very similar for all three policy-simulations.

An earned income tax credit is introduced in a dynamic CGE model with overlapping generations and imperfectly competitive labor market segments. The model is calibrated to a 10 percent sample of a register of all persons and workplaces in Denmark and scaled to the level of National Accounts. Each labor market segment is covered by a collective bargaining agreement, except one that is competitive. In all segments a distribution of workers with different productivity exists. A tax credit, which increases the net income of the poorest full.time employed workers in the economy by DKK 5000, reduces unemployment by 2.5 percentage points. Since the marginally employed have a productivity lower than the average worker, employment measured in productivity units increases only by approximately 1.5 percentage points. Women and younger generations tend to gain from the policy. The effect of the policy depends crucially on the initial level of unemployment because the productivity distributions are highly non-linear: if the initial level of unempolyment is halved to 5 percent so are the effects of unemployment and employment.

The structural effects of the Danish tax reform of 1993 are evaluated using DREAM, a dynamic CGE model of the Danish economy featuring overlapping generations of agents who have perfect foresight. The tax reform implied a reduction in the tax burden and the progressivity of the labor income taxation, an introduction of "green" taxes as revenue raising instrument and a restructuring of the capital income taxation. We find that the overall macroeconomic effects of the total reform are limited, but that the accumulation of wealth in the private sector is stimulated, which generates a long run increase in aggregate consumption. Analyzing the three parts of the reform separately reveals that the small net effect of the reform is due to counteracting forces of each of the parts. The reform is a strict Pareto improvement - given the initial calibration of the model - in the sense that all generations are better off after the reform.

To a large extent, Danish wage contracts are minimum wage contracts, where the wage of an employed worker consists of a negotiated minimum wage and a personal raise. This paper considers bargaining over minimum wages given that the members of the union have different levels of productivity. We show that the total wage of an individual in this case depends on both individual productivity and standard opportunity costs of the union. Further, it is shown that ceteris paribus the minimum wage is lower in this economy than the (total) wage in an economy where all workers have the same productivity. Finally, it is shown that increasing progressivity of the tax system does not necessarily imply a reduction in the minimum wage. A log-linearized version of the wage equation is estimated for 6 major unions using a panel which is a sub-sample of a representative 5 per cent sample of the Danish population. Estimated parameters of the marginal tax are negative for most male workers whereas this is not the case for women. Even for male workers the coefficients are much smaller than in previous studies based on macro data.

This paper gives an introduction to computable general equilibrium (CGE) models. The paper focuses on static CGE-models - how they are constructed, calibrated and used for policy evaluation. The paper also briefly presents the handful of models that at present are being used or are under development in Denmark. It is also discussed how the method differs from the traditional macroecenometric models, as well as the pros and cons of the approach.

Conventional generational accounts are insufficient as instruments to judge the appropriateness of policies which may change intergenerational distribution. One of the reasons is that they are forward-looking and hence do not measure the impact of historical events. We construct a measure of full lifetime incomes and net government contributions for all generations born in Denmark from 1930 to 2030, combining historical figures with future projections on the large computable general equilibrium model DREAM. Despite many short-term fluctuations, the resulting development in lifetime income figures is relatively smooth, and net government contributions make up only a modest part of each generation’s consumption possibilities, even though present and expected future tax rates in Denmark make up no less than half of GDP.

This paper presents a quantification of the economic impacts to Denmark of the European Unions’ upcoming East enlargement. Specifically the effects of customs liberalization, market integration and immigration are quantified with due regard to estimated repercussions to transfers to and from the European Union and enhanced growth in the new member states. In the basecase projection, which excludes additional immigration, a steady state increase of aggregate GDP at factor costs of 0.08 percent is found along with a welfare gain of 1.27 percent of 1998 GDP. The welfare gains are demonstrated to be highly dependent on the catching up of the CEE region and the route taken by the EU for funding the enlargement.

In a world where labor earnings are uncertain and borrowing-constraints are present, progressive taxation is likely to have risk mitigating benefits for consumption-smoothing agents; higher tax payments are due in periods of life when income is relatively high, and less tax must be paid when income is low. This lowers the probability that the borrowing constraint becomes binding. The question is if this income-smoothing risk-mitigating property of progressive taxation has any value to the consumers? Simulations using a large-scale computable general equilibrium model with competitive markets show that consumers prefer progressive taxation of labor earnings to proportional taxation - a result that is contrary to the findings in the deterministic framework by Auerbach and Kotlikoff (1987). However, there is a trade-off between the positive risk-migating properties of progression, and the negative distortionary effects; indeed it turns out that there is an optimal level of progressivity.

This paper examines the consequences of introducing an idiosyncratic uncertain interest rate in a standard life-cycle model à la Auerbach and Kotlikoff (1987). Since the labor market has no uncertainty, labor earnings are used by the consumers to compensate for the risks in the capital market. The multi-period general equilibrium model introduces the possibility for consumers to adjust their labor supply ex post in response to new information becoming available (in addition to the opportunity to hedge ex ante). Increased uncertainty causes the number of hours worked to increase, since some old agents start supplying labor to compensate the poor performance of their savings. The framework also makes it possible to quantify the value of labor supply flexibility for these old agents.

This paper investigates the importance of the usual assumption of divisibility. In the labor market a finite set of choices is introduced: between working full or part-time or not to work at all. To add realism and to ensure smooth aggregate behavior the option of limited overtime for individuals working full time is introduced. The simulations show that indeed indivisibilities matter - the results obtained in each of the two models are markedly different. The impact of the policy experiment (a move from progressive to proportional taxation of labor income) is much larger in the case where the labor supply is continuous; the welfare gains of the switch from progressive to proportional taxation is almost 150 percent larger with continuous labor supply. The sensitivity analysis shows that this result depends on how the indivisibilities are specified, but in almost all cases are the welfare gains from the tax reform more than twice as large in the continuous model.

Recently David Laibson and others have argued in favor of using hyperbolic discount functions. The purpose of this paper is to investigate whether conventional wisdom, based on the standard model with exponential discounting, also holds in the case where consumers have hyperbolic discount functions. In other words do hyper-bolic preferences matter for practical policy evaluation? Within the framework of a suitably modified standard General Equilibrium model à la Auerbach and Kotlikoff, this is done by simulations of both fundamental changes in the tax base, as well as more marginal experiments comparing the excess burden of taxation. Based on the simulations it turns out that the answer to the question is a maybe: if preferences are sufficiently hyperbolic then policy conclusions change. Unfortunately this degree of hyperbolicness in the discounting function is at the level that is considered realistic by empirical studies.

The efficiency gains of VAT, income tax, corporate tax and capital income tax reductions are evaluated in the dynamic CGE model DREAM. The tax rates are reduced from 5% to 25%; the reduction is fully financed by lump sum taxation. The efficiency gains are calculated as the marginal and average efficiency gains defined as change in society welfare measured in kroner per kroner change in tax revenue. We find that a decrease of 5% results in marginal efficiency gains in the area of 9.49% to 22.94%. The reduction of the VAT increases the society welfare due to intertemporal redistribution of the households’ welfare. In the case of an income tax cut the society gain welfare because almost all-living and future households are better off. The reduction of the capital income result in a society welfare loss since both living and future households suffers welfare loss. However, the marginal efficiency gain is still positive since the reduction causes a tax revenue gain. In spite of the existence of the ” tax paradox” the society gain welfare by a reduction of the corporate tax. We argue that the chosen financing rule and the discount rate in the society welfare function heavily affects the measures of efficiency as well as the individual household’s welfare.

Two macroeconomic models for Denmark; the macroeconometric model ADAM (Annual Danish Aggregate Model) and the dynamic CGE-model DREAM (Danish Rational Economic Agents Model) are compared. An analytical model that contains as special cases a representation of the stationary state in each of the two large simulation models is set up. Analytical multipliers are compared for three different types of policy. In simulations the entire dynamic path of the marginal effects of the same policy changes are compared. Our conclusions are: 1) The qualitative long run effects are similar in the two models. 2) The quantitative effects are in most cases comparable in the long run. 3) The dynamic evolutions of the multipliers are surprisingly similar. A major contributor to quantitative differences is the dissimilarity of the wage curves in the two models: A negative supply effect is the outcome of an increase in the tax burden in DREAM whereas no supply effect is present in ADAM. Second, the price-wage mechanism in ADAM is much stronger than in DREAM. Even if the modelling of private consumption differs significantly between the models, the evolutions of total private consumption in the two models are very similar for all three policy-simulations.

An earned income tax credit is introduced in a dynamic CGE model with overlapping generations and imperfectly competitive labor market segments. The model is calibrated to a 10 percent sample of a register of all persons and workplaces in Denmark and scaled to the level of National Accounts. Each labor market segment is covered by a collective bargaining agreement, except one that is competitive. In all segments a distribution of workers with different productivity exists. A tax credit, which increases the net income of the poorest full.time employed workers in the economy by DKK 5000, reduces unemployment by 2.5 percentage points. Since the marginally employed have a productivity lower than the average worker, employment measured in productivity units increases only by approximately 1.5 percentage points. Women and younger generations tend to gain from the policy. The effect of the policy depends crucially on the initial level of unemployment because the productivity distributions are highly non-linear: if the initial level of unempolyment is halved to 5 percent so are the effects of unemployment and employment.

The structural effects of the Danish tax reform of 1993 are evaluated using DREAM, a dynamic CGE model of the Danish economy featuring overlapping generations of agents who have perfect foresight. The tax reform implied a reduction in the tax burden and the progressivity of the labor income taxation, an introduction of "green" taxes as revenue raising instrument and a restructuring of the capital income taxation. We find that the overall macroeconomic effects of the total reform are limited, but that the accumulation of wealth in the private sector is stimulated, which generates a long run increase in aggregate consumption. Analyzing the three parts of the reform separately reveals that the small net effect of the reform is due to counteracting forces of each of the parts. The reform is a strict Pareto improvement - given the initial calibration of the model - in the sense that all generations are better off after the reform.

To a large extent, Danish wage contracts are minimum wage contracts, where the wage of an employed worker consists of a negotiated minimum wage and a personal raise. This paper considers bargaining over minimum wages given that the members of the union have different levels of productivity. We show that the total wage of an individual in this case depends on both individual productivity and standard opportunity costs of the union. Further, it is shown that ceteris paribus the minimum wage is lower in this economy than the (total) wage in an economy where all workers have the same productivity. Finally, it is shown that increasing progressivity of the tax system does not necessarily imply a reduction in the minimum wage. A log-linearized version of the wage equation is estimated for 6 major unions using a panel which is a sub-sample of a representative 5 per cent sample of the Danish population. Estimated parameters of the marginal tax are negative for most male workers whereas this is not the case for women. Even for male workers the coefficients are much smaller than in previous studies based on macro data.

This paper gives an introduction to computable general equilibrium (CGE) models. The paper focuses on static CGE-models - how they are constructed, calibrated and used for policy evaluation. The paper also briefly presents the handful of models that at present are being used or are under development in Denmark. It is also discussed how the method differs from the traditional macroecenometric models, as well as the pros and cons of the approach.

This paper gives an introduction to the Computable General Equilibrium (CGE) modelling, and presents an application of the technique to the analysis of the Europe Agreements between the EU and Hungary, Poland and the former Czechoslovakia. The main purpose of the paper is to illustrate the method, rather than present a state-of-the-art analysis. The CGE-approach makes it possible to pursue the analysis further than possible with analytical methods, and it can yield qualitative as well as quantitative results. A model is presented, and it is used to analyse the consequences of the Europe Agreements as well as the sensitivity of the results to important assumptions. The analysis shows only moderate long run gains for the Eastern Europe countries (around 1-2% of GDP per year), and very small gains for the EU countries. The sensitivity analysis shows that the results are relatively robust to the way the model is calibrated.