7. METODOLOGÍA
7.3 COMPARACIÓN DE DOS ESTRATEGIAS DE PREVENCIÓN Y CONTROL DE LA
Empirical studies of work incentives generally measure the financial incentive facing individuals in the form of replacement rates, the ratio of income when unemployed to income when in work. Our analysis of replacement rates in Ireland via the SWITCH microsimulation model is based on data obtained in a large-scale household survey carried out in 1994 as part of the Living in Ireland survey series. The data is then uprated annually to reflect growth in earnings and reweighted to reflect changes in important population parameters e.g. the unemployment rate. (See Callan et al., (2001) for a full description of the uprating process.). At the same time, the annual tax and social welfare policy changes (tax rates and bands, social welfare rates etc.) are included in the microsimulation model.
The financial incentive to work for the unemployed depends on the expected wage. This cannot be observed directly, but is modelled econometrically based on the age and education levels of the individual. The net impact of additional earnings on the family unit can then be simulated using the tax-benefit model. This takes account not only of tax and social insurance deductions from pay, but also of reductions in means-tested benefits, including the Family Income Supplement scheme. For a fuller description, see Callan et al. (1997). There has been one methodological development since then which is worthy of note in this context. The counterfactual wage now includes a random error term, which makes the proportion of very low and very high wages closer to what is found in reality, and as a result, gives rise to a greater incidence of high replacement rates. The simulation of out-of-work incomes of those currently employed is based on predicted welfare entitlement versus actual in-work income. 7
Distribution of replacement rates in 2002
Table 2.5 shows the distribution of replacement rates for 2002 according to three main types of labour force activity. The table shows that the numbers of individuals currently employed who would be better off out of work (i.e. with a replacement rate of 100 per cent of more) is negligible, while about 4 per cent of those currently unemployed face such replacement rates. It is assumed that two-thirds of those entitled to a Family Income Supplement payment actually take it up..8 About one in six of those unemployed and in receipt of unemployment payments (benefit or assistance) face a replacement rate of over 70 per cent.
Table 2.5 Distribution of replacement rates in 2002 across activity
Employee Unemployed and in
receipt of Unemployment Assistance or Benefit Home Duties <40% 49.8 49.8 9.6 <50% 17.2 17.3 14.6 <60% 13.8 11.9 15.7 <70% 9.0 6.9 24.0 <80% 5.1 6.5 20.8 <90% 2.8 4.0 12.1 <100% 1.4 2.4 3.1 >100% 0.8 3.7 0.3 Total 100.0 100.0 100.0 1,261,900 59,200 510,900
A low replacement rate for someone currently employed indicates that they would be significantly poorer if they were to become unemployed. This shows that for over 600,000 individuals the financial incentive to remain in employment is very strong (as their replacement rate would be 40 per cent or lower). The number of people engaged in home duties is often considered to be an element of the potential labour supply that can be particularly responsive to financial incentives to work. Our estimates suggest that over one-third of these individuals face a replacement rate of more than 70 per cent in 2002.
7
Non-cash benefits such as the value of medical card entitlement, fuel allowance and differential rent for local authority tenants are not taken into account in these calculations.
8
The allocation of individuals to take-up or non-take-up is random. Given that full take-up of FIS has never occurred, it is not possible to assume all who are eligible will benefit from this income supplement in a counterfactual scenario.
Table 2.6 Distribution of Replacement rates for unemployed persons in receipt of UA or UB 1994 1998 2002 <40% 24.3 39.4 49.8 <50% 18.0 19.0 17.3 <60% 14.7 13.4 11.9 <70% 13.1 10.3 6.9 <80% 12.1 7.9 6.5 <90% 5.9 5.3 4.0 <100% 6.6 3.1 2.4 >100% 12.8 4.8 3.7 Total 100.0 100.0 100.0 198,200 137,200 59,200
The proportion of all those in receipt of UA or UB with a high replacement rate of 70 per cent or above halved between 1994 and 2002 falling from 30.8 per cent in 1994 to 14.1 per cent at 2002. The falls in unemployment shown in Table 2.6 are due to both structural change in the population between the years shown as well as the effect of tax and social welfare policy changes.
Tax and welfare policy impact
In order to assess the impact of policy changes, we must ask first of all what would constitute a “no policy change” scenario. One interpretation, commonly used in the construction of the annual budget, is that tax and welfare policies would be frozen in nominal terms. We term this the “conventional opening budget”. This is at odds, however, with the approach used in framing much of the rest of the budget, when the concept of “no policy change” is translated into “constant levels of service”. This could equate to price indexation. But on a broader view, constant levels of welfare service could be seen as linked to indexation in line with wages or earnings.
How should the choice between these different frames of reference be made? In what follows, we provide information on policy impacts assessed against each of the three benchmarks, since readers may be more familiar with only one of them. But in our view it is the wage-indexed budgetary policy which provides the most appropriate benchmark against which to assess actual policy changes. With rising real incomes, the share of income taken in taxes will rise under either the conventional opening budget, or the price-indexed budget. Only under the wage-indexed budget will the share stay constant. From a distributional point of view, only this benchmark would ensure that similar growth in net income was enjoyed by different income groups.
The implementation of this approach is straightforward. Tax rates are held constant, but tax credits or tax allowances, and income tax bands, are indexed either in line with prices or with wage growth. Similarly, welfare rates are indexed either in line with wage growth or with price inflation. Our estimate of economy-wide wage growth between 1998 and 2002 is 29.8 per cent, while price growth is estimated at 16.8 per cent.
Table 2.7: Policy impact from 1998 to 2002 (UA/UB recipients)
Price indexed budget
1998-2002 (neutral) budget Wage- indexed 1998-2002
Actual 2002 policy
Index 16.8% 29.8% Not applicable
<40% 51.2 50.7 44.6 <50% 17.2 18.4 19.4 <60% 12.0 15.3 12.6 <70% 7.8 5.4 7.8 <80% 5.7 4.9 6.8 <90% 3.9 3.1 3.8 <100% 0.8 1.9 3.5 >100% 1.5 0.2 1.5
Total 100.0 100.0 100.0
59,200 59,200 59,200
Table 2.7 shows that the proportion facing a low replacement rate (less than 40 per cent) has fallen compared with both the price- and wage-indexed benchmarks. In terms of the broader distribution of replacement rates, the policy impact measured relative to either the price-indexed or wage-indexed benchmark involves an upward shift in replacement rates.
Impact of the Minimum Wage on Replacement Rates
Callan and Walsh (2000) estimated the impact of the introduction of the national minimum wage on the distribution of replacement rates for the unemployed. Those results, reported in Table 2.8, show a substantial downward shift in replacement rates associated with the introduction of a national minimum wage equivalent to Ir£ 4.40 in 2000.
Table 2.8: Impact of the Introduction of a National Minimum Wage, 2000 (UA/UB recipients)
Baseline 2000 2000 with National
Minimum Wage <30% 19 20 30-40% 22 22 40-50% 19 36 50-60% 14 6 60-70% 9 6 70-80% 7 5 80-90% 5 4 90-100% 4 1 >100% 2 1 Total 100 100
Since then, however, wage growth in the wider economy has been more rapid than the increase in the National Minimum Wage, which will stand at the equivalent of Ir£ 4.70 for most of the year 2002. By contrasting a scenario based on current, 2002 policy with a hypothetical 2002 policy without a National Minimum Wage we can identify the current impact of the national minimum wage on the distribution of replacement rates. This is shown in Table 2.9. It is clear that the impact of the National Minimum Wage in 2002 is a good deal more limited than in 2000, but is still significant.
Table 2.9: Impact of the Introduction of a National Minimum Wage, 2000 (UA/UB recipients)
Baseline 2002 2002 with National
Minimum Wage <30% 27.9 28.2 30-40% 21.9 22.5 40-50% 17.3 18.4 50-60% 12 15.3 60-70% 6.7 5.4 70-80% 6.5 4.9 80-90% 4.1 3.1 90-100% 2.3 1.9 >100% 1.3 0.2 Total 100 100