Development in the Labour Supply

Erik Haller Pedersen and Johanne Dinesen Riishøj, Economics

INTRODUCTION AND SUMMARY

The labour force has increased during the most recent upswing, even though demographic trends would warrant a reduction. Good job op portunities have led to an expansion of the labour force and moved more people off public benefits, and at the same time there has been an influx of foreign labour. Particularly the extent to which Denmark has succeeded in attracting foreign labour distinguishes this upswing from previous ones. In addition, the total number of hours worked per em ployee has stopped declining and has in fact been on the rise since the mid-1990s. Together, the increase in the labour force and the higher number of hours worked per employee have boosted the labour supply by more than 160,000 since 2004.

Employment has risen even more than the labour force, and registered unemployment has fallen to the lowest level since the early 1970s. Unemployment can be expected to increase as the economy cools down. Nevertheless, the challenge of the coming years will be to increase age-related participation rates and to ensure a sufficient supply of foreign labour as the ageing of the population continues to reduce the labour supply significantly. The labour supply is only expected to increase again in 10 years' time as welfare reform initiatives gradually raise the eligible ages for early retirement benefits and state retirement pension as from 2019 and 2024, respectively. Above all, an increase in the labour supply is necessary to ensure fiscal sustainability.

In Denmark a very large share of the population of working age is outside the labour force and on some kind of public benefits. The share is the highest among the Nordic countries. An increase in the age-related participation rates would therefore require a reduction of the number of people on public benefits.

The requirement for an increase in the labour supply to maintain the level of output and ensure the sustainability of public finances and ulti mately the foundation of the Danish welfare society will be so urgent in the coming years that a wide range of initiatives are needed.¹

LABOUR FORCE AND LABOUR SUPPLY

The labour force is calculated as the sum of those in employment and the unemployed, expressed in number of people. Employment increased more than the supply of labour during the most recent upswing and resulted in a substantial reduction in unemployment, cf. Chart 1.

The total labour supply is determined by the labour force and by the hours worked per employee. Thus, the labour supply can be increased both by attracting more Danes and foreigners to the labour market and by raising the number of hours worked by those in employment.

The labour supply is affected by cyclical changes. In times with good job opportunities more people will enter the Danish job market while the opposite is true in bad times. The labour supply is also affected by structural factors such as demographic trends, social norms, rules on eli gibility for unemployment benefits, leave-of-absence and retirement schemes, e.g. early retirement benefits, access to social pensions, and the official retirement age.

Relating the labour force to the population of working age gives the average participation rate. Breakdown by age intervals gives the age-related participation rates, cf. Chart 2. As people close to retirement age are less active on the labour market than the middle-aged, a shift in the population towards more elderly people would lead to a reduction of the average participation rate, even if the age-related participation rates remain unchanged. The average participation rate will also fall to the extent that immigration to Denmark consists of people with a par ticipation rate below that of the residents.

Demographic trends will tend to lower the average participation rate in the coming decades. At the same time the share of the population outside the labour market will increase.

Cyclical effect on the labour supply
As already mentioned, the labour supply is determined by both the economic cycle and underlying structural factors. To get an indication of the cyclical effect on the labour supply, a gross participation rate is com pared with a trend in Box 1. In addition to those in employment and the unemployed, the gross participation rate also includes people on early retirement benefits and transitional allowance, people on maternity leave, child-minding or educational leave and people participating in educational labour-market schemes as a share of the population aged 15-74 years, thus isolating the cyclical effect from the main policy-determined structural effects.

CYCLICAL EFFECT ON THE LABOUR SUPPLY	Box 1
The labour force equals the sum of those in employment and the unemployed. The participation rate shows the relation between the labour force and the population of working age (15-74 years). In order to get an idea of the impact of cyclical factors on the labour supply, a gross participation rate is calculated, as well as a trend expressing the effect of structural conditions. The gross participation rate is calculated using a gross labour force, including people on maternity leave, other leave and transitional allowance, people participating in educational labour-market schemes and early retirement benefit recipients aged 60-64 years, thus adjusting for a large number of policy-determined changes in the labour force.
GROSS PARTICIPATION RATE AND TREND	Chart 3
1The gross participation rate calculation is based on a gross labour force including, in addition to those in employment and the unemployed, people on early retirement benefits, transitional allowance, maternity leave, child-minding and educational leave as well as people participating in educational labour-market schemes. The logistic trend is in accordance with the MONA model. 2Statistics Denmark and own calculations.
The gross participation rate is compared with a logistic trend with its saturation point in the 1990s. From 1998, the trend variable changes to fall in step with a standard labour force calculation, cf. Chart 3, thus adjusting for demographic trends in the last part of the period. The logistic trend is a method to smooth out the gross participation rate. It was chosen because of the s shape's ability to capture women's increased participation in the labour market from the 1960s to the 1990s, when women's participation rate had reached a level virtually on a par with men's. Women's increased participation in the labour market completely dominates demographic trends in the first decade of the period under review. The cyclical contribution to the labour force appears as the difference between the gross participation rate and the smoothed-out curve, i.e. as the difference between the two graphs in the Chart.

The spread between the gross participation rate and the participation rate increased from the end of the 1970s to the mid-1990s following the introduction of several schemes such as early retirement benefits, transi tional allowance, leave, etc. With 140,000 people, early retirement bene fits is the most significant of these schemes. The spread has stabilised since then, cf. Chart 4, and indicates a current labour force potential of up to 6 per cent of the population in the age groups concerned.

The cyclical contribution, calculated as the difference between the gross participation rate and the trend, can be compared with the output gap, which is a measure of the pressure on the economy, cf. Chart 5. Historically, there has been a clear correlation, in that increased pres sures on the economy have led to both reduced unemployment and an expansion of the labour force. During the most recent upswing the cyclical effect seems to have been more pronounced than previously. The increase has been even stronger than shown in the Chart, since the national accounts do not include in the labour force employees se conded to Denmark from foreign companies.

The challenge of the coming years will be to prevent the participation rates from falling back again as the economy slows down, but rather to ensure a further increase and an influx of foreign labour. It is important to turn what was initially a cyclical effect into a more permanent in crease in the labour force.

The increase in the labour force during the boom economy was to some extent the result of fewer young people enrolled in education, however. This may reflect that fewer young people choose to get an education, but also that they finish their studies sooner. This is a well-known phenomenon in a boom economy. If fewer young people get an education, the level of competence will be undermined, and therefore such expansion of the labour force is not sustainable in the long run. On the contrary, a higher proportion of young people need to get a quali fying education if Denmark is to sustain a long-term level of prosperity and competitiveness.

Not only has the labour supply grown in numerical terms during the upswing, the number of hours worked per employee has also increased, cf. Chart 6. The increase in working hours per employee from 2004 to 2007 corresponds to 34,000 more people in full-time employment. This is a continuation of the trend that has been observed since the mid-1990s with an increased number of working hours per person. In this respect, Denmark differs from practically all other OECD countries, where the number of working hours continues to fall. Denmark remains at the lower end among the OECD countries, however, which is a reflection of its high participation rate.

Where does the influx to the labour force come from?
The labour force, calculated as the sum of those in employment and the unemployed in the 15-64 age group, grew by 68,000 from 2004 to 2007. If the underlying demographic trend with fewer people of working age is taken into account, there has actually been an influx of labour of 81,000, cf. Table 1.

In Denmark virtually everyone outside the labour force is on some kind of public benefits. The increase in the labour force includes 30,000 that previously received public benefits. Among the cash-benefit recipients, the proportion of people in match groups 4 and 5 keeps growing as those with better qualifications find jobs. This makes it increasingly difficult to source additional labour from the group of cash-benefit recipients. The number of people on sickness benefits has risen during the period under review.² The reason is that as employment rises, weaker groups with higher rates of absence due to illness have entered the labour market.

Foreign labour accounted for approximately 102,000 additional em ployees during the boom, thus contributing more to meeting the need for labour than during previous upswings.

The number of employees seconded to Denmark is subject to some un certainty, cf. Box 2, and is not included in the calculation of the labour force in the national accounts. The increase in the participation rate during the most recent upswing is therefore even stronger than indi cated in Chart 4.

FOREIGN LABOUR STATISTICS	Box 2
Complete statistics are not available regarding foreign labour in Denmark, cf. Chart 7, and to some extent the figures relating to foreign labour in this article are therefore based on estimates. All employees in Denmark except those seconded are included in the employment calculation in the national accounts and in the ATP (Labour Market Supplementary Pension) statistics, but they are not broken down by nationality in these or other economic statistics regarding employment. Employment is broken down by nationality in the register-based labour force, but this only includes people residing in Denmark, and as a result the statistics do not include frontier workers and people seconded to Denmark. Furthermore, the register-based labour force is only compiled on an annual basis and with a considerable lag. Consequently, other sources must also be applied when assessing the contribution of foreign labour to the labour force.
FOREIGN LABOUR IN THE DANISH LABOUR MARKET	Chart 7
Note: Year-end observations. The 2008 observations are from August. It should be noted that there are no fig ures for the number of German frontier workers in 2004, so the 2004 figure shows only Swedish frontier workers, whereas the 2005, 2006, 2007 and 2008 figures include both Swedish and German frontier workers. The frontier workers include only people who were not born in Denmark, i.e. Danes residing in Sweden or Germany are not included. Source: Tax Øresund, Tax Centre Tønder, the Danish Immigration Service, Statistics Denmark and own calculations.
The estimated number of immigrants in employment from western and non-western countries, respectively, is based on the register-based labour force statistics. The statistics are compiled in November, but Statistics Denmark places the observation in January of the following year. The November observation is included as year-end in the present estimate. As the most recent statistics are from November 2006, the figures have been projected using the number of immigrants in age group 16-66 (from western and non-western countries, respectively), and the employment rate for immigrants (western and non-western, respectively) aged 16-66 years. The calculation maintains the employment rate of November 2006. As no data is available on the total number of frontier workers in Denmark we have assumed that they only include employees from Germany and Sweden. Frontier workers residing in Sweden and working in Denmark are registered in the Øresund database, and fairly exact data on them is therefore available. The most recently published figures are from November 2006. The subsequent projection is based on the granting of tax deduction cards to new Danish taxpayers residing in Sweden. The number of German frontier workers is based on an estimate from Tax Centre Tønder. The calculation of employees seconded to Denmark is also based on information from Tax Centre Tønder. Information on the number of foreign companies conducting business in Denmark without their representatives residing permanently in Denmark is available from SKAT. The number of employees seconded to Denmark is based on the estimate that each company employs an average of 15 people. The estimate is based on a sample survey, cf. Jens A. Hansen and Søren Kaj Andersen, Eastern European Workers in the Danish Construction Sector (in Danish only), FAOS (Employment Relations Research Centre), January 2008, and subject to uncertainty.

If the increase in working hours per employee is included, the total in flux to the labour market amounted to 166,000 from 2004 to 2007.

The boom in the economy since 2003 has influenced the share of people aged 60-64 who receive early retirement benefits only to a small extent. There has been a decline at the lower end of the age interval and a flat development at the upper end, cf. Chart 8.

If the share of early retirement benefit recipients had been the same in 2007 as in 2004, 8,000 more people would have been on early retire ment benefits.

The Danish early retirement benefit scheme contributes to explaining why the share of people aged 16-66 on public benefits is the highest among the Nordic countries, cf. Chart 9.

THE LABOUR FORCE IN A FORWARD-LOOKING PERSPECTIVE

For many years population trends were favourable for Denmark in that the population groups with the highest participation rates increased. This has changed. Up to 2020 the labour force will grow by 75,000 in the 15-29 age group, but if the participation rates remain the same as in 2005, the number of people aged 30-54 will fall by approximately 140,000 during the same period, cf. Chart 10. The assumption of an un changed participation rate for young people may be too optimistic given the objective that an increasing share of young people in each year group should get a qualifying education.

If the age-related participation rates and hours worked per employee are kept at the 2005 level, the total hours worked will fall by 3.5 per cent up to 2020. At the same time, the number of elderly people and children will increase by a total of 240,000, thereby draining public funds by an additional estimated kr. 13 billion³ per year at present value. The need for teachers for the children and for nursing and healthcare professionals for especially the elderly will increase and has to be met by a reduced labour supply. Higher expenditure on public service combined with a smaller tax base increases the urgency of the fiscal sustainability issue.

While population trends can be calculated fairly accurately for a con siderable number of years to come and are difficult to manage politic ally, that is not the case for labour participation and hours worked per person per year. These elements can be influenced by politically adopted incentive mechanisms and amendments to the framework conditions. Experience from Danish labour-market policies over the last 15 years indicates that a "stick and carrot" approach has an effect on labour in put.⁴ If the Danish welfare model is to be maintained without substan tial reductions, a wide range of measures will be needed to increase labour input in the economy.