Key points

There was a construction boom in New Zealand between 2001 and 2008. The main finding of this paper is that the construction industry^[1] found a great many additional workers to meet the rapidly rising demand without very large increases in labour costs.

Menaka Saravanaperumal in Labour Market Adjustment in the Construction Industry 2001–2006, (Statistics New Zealand, 2008) wrote on construction labour market adjustment using the Linked Employer-Employee Dataset (LEED)^[2]. Updating these results to the year to March 2009, the latest available data, the key findings are:

• There was a 54% increase in employment in the construction industry between 2001 and 2008, from 125,000 in 2001 to 194,000 in 2008, followed by a 3.5% fall in the year to March 2009. This was compared with a 23% increase in employment across all industries between 2001 and 2008.
• The majority of inflows into the construction industry came from alternative industries, with about six of every 10 new construction workers having worked in another industry in the previous year. Most of the rest were recruited from outside of the labour force rather than from being on a benefit.
• Turnover rates within the construction industry were not unusually high and tended to fall between 2001 and 2009.
• Pay increases in construction from 2001 to 2009 were only marginally higher than those across all industries across a number of wage growth measures.

The Labour Cost Index (LCI), which adjusts for human capital and hours of work supplied, shows that wages in construction increased by 3.0% per year over this period compared with 2.8% for all industries. In what was a strong economy of the mid-2000s, the construction industry recruited many more workers without a relatively large increase in real labour costs.

Labour demand in the construction industry

As the economy slowed in recent years, employment in the construction industry fell after the surge in the mid-2000s (see Figure 1). Employment increased from 125,000 in the year to March 2001 to quickly peak at 194,000 in the year to March 2008, only to slowly taper to 187,000 in March 2009. The construction industry was able to draw in these many new workers in a short time without having to raise wages by a large amount relative to other industries. There was a 78% increase in the value of building consents between 2002 and 2008, which then fell by 14% between 2008 and 2009 (see Figure 2).

Figure 1: Employment in the construction industry (ANSIC06), 2001-2009

Data table for Figure 1

Figure 2: Value of building consents issued, total building and construction, 1991-2009

Data table for Figure 2

Inflows to construction

The number of workers flowing into and out of a labour market each year and over the business cycle is an indicator of its ability to cope with rapid changes in demand in the future. For example, increases in the labour supply in some rapidly expanding labour markets can be slow because there can be substantial initial specialised education and training requirements for new recruits. Other industries can find additional workers faster because there are pools of suitably qualified recruits with more transferable general and occupational skills who are currently either working in other industries or are out of the labour force.

As shown in Figure 3, the LEED data suggests that net inflows of workers from other industries into the construction industry rose rapidly from 768 people in 2001 to 9,723 people in 2006, then fell afterwards, reaching a net out flow of 93 people in 2009. In most years, inflows from other industries were the largest source of total net inflows.

Figure 3: Net inflows of into the construction industry, 2001-2009

Data table for Figure 3

After 2005 the growth of inflow slowed substantially, in line with the slowdown in employment growth, but did not reach a net outflow in total until 2009 (see Figure 3). There appears to be a large pool of workers not currently working in the construction industry who can move into construction jobs when demand picks up.

Most of the increases in inflows of workers into construction were due to a rise in inflows rather than fewer exits from construction. The construction industry tends to meet demand for additional labour through recruitment rather than through increased retention. A steady 26,000 to 33,000 workers left construction in each year in the boom years between 2001 and 2006, increasing to over 42,000 in 2009 as the economy and the construction industry slowed. The number of entrants increased from 29,000 in 2001 to 46,000 in 2006, before falling back to just over 36,000 in 2009 (see Figure 4).

Figure 4: Gross inflows of workers from other industries and gross outflows of workers to other industries from the construction industry, 2001-2009

Data table for Figure 4

Figure 5 shows that inflows from outside the workforce doubled while inflows from other industries increased by 63% between 2001 and 2006, with both dropping by around 20% between 2006 and 2009. Individuals who were previously out of the labour force includes those out of the tax system such as students, migrants, and those undertaking caring responsibilities; and those receiving ACC, government superannuation, paid parental leave, or student allowance as their main income source in the previous year. Inflows from income-tested benefits include those receiving an income-tested main benefit^[3] as their main income source the previous year.

Figure 5: Gross labour inflows for the construction industry, by source, 2001-2009

Data table for Figure 5

The number of former benefit recipients that entered the construction industry was fairly steady between 2001 and 2005, but fell by 62% between 2005 and 2009. The number of construction workers who were previously on the benefit never exceeded 5,000 in any year between 2001 and 2009.

Inflows from other industries accounted for the largest proportion of total inflows of workers. In 2009, over 64.1% of the total inflows of workers into construction from other industries, particularly manufacturing (17.8%), administrative and support services (12.0%), and retail trade (11.5%). While manufacturing was consistently the largest source of inflows from other industries, Figure 6 shows that the sources of these inflows of workers were widely spread across industries.

Figure 6: Labour inflows into construction by detailed source, 2009

Data table for Figure 6

Outflows from construction

Outflows to other industries, out of the workforce, and on to benefits all fell between 2001 and 2003 (see Figure 7). From 2004 to 2009, outflows to other industries and to out of the workforce increased in each year, and outflows to benefits remained fairly steady. The outflows to out of the workforce are likely to reflect an increasing number of retirements as a result of the ageing workforce. The share of outflows to out of the workforce who were aged 55 years and over increased from 13% in 2001 to 18% in 2009. It could also reflect an increase in emigration of construction workers.

Figure 7: Outflows from the construction industry, by destination, 2001-2009

Data table for Figure 7

Industry and worker turnover in the construction industry

In addition to the large flows in and out of the construction industry each year, there are also many people moving between jobs within the construction industry. Table 1 shows that industry level turnover, defined as the average of outflows and inflows during the year divided by total employment in the construction industry, fell slightly between 2001 and 2009.

Source: Statistics New Zealand, LEED annual data. March years.

Note: Worker turnover is published by Statistics New Zealand as the annual average quarterly rate, which had been multiplied by four in this report to provide an annual figure.

Worker turnover, which includes people moving between jobs within the construction industry, has been much more variable, falling by ten percentage points between 2001 and 2009, with a particularly large drop in 2009. Worker turnover, defined as the average of the number of job starts and job finishes divided by the average number of jobs during the year, is not directly comparable to industry turnover as it excludes the self-employed, who made up 33-40% of construction employment over this period. However, the figures indicate that the amount of churn within the industry is substantially higher than the movement in and out of construction. It is also important to note that worker turnover in the construction industry is not unusually high compared to other industries.

Relative wage increases in the construction industry

Increases in relative wages are an important indicator of the ability of a labour market to adjust to increases in demand. This is also known as the elasticity of labour supply. A highly elastic labour supply indicates that new workers can easily be brought into the labour market as new jobs arise, without a large increase in the prevailing wage rate. In a highly inelastic market, for example where long pre-employment training is required, an increase in labour demand would exert much more upward pressure on wages. Labour supply will increase in time as these newly trained recruits enter the market in response to the increased rewards for risking a relatively large upfront investment in skills.

In this report we have seen evidence of a very strong increase in demand in the construction industry from 2001 to 2006, and a corresponding inflow of workers from other industries and from outside of the labour workforce, suggesting an elastic labour supply. However, this was also a period of strong demand for skills and labour across the entire New Zealand economy, which might indicate that the rapid increases in construction labour demand should lead to relative wage increases.

Three datasets were used to investigate the responsiveness of wages in the construction industry during the period of heightened labour demand. These include Linked Employer-Employee Dataset (LEED), the Quarterly Employment Survey (QES) and the Labour Cost Index (LCI). LEED measures the average quarterly earnings paid to full-quarter employees in the construction industry^[4]. Full-quarter earnings can be split into earnings for continuing jobs and new hires^[5]. LEED does not distinguish between part-time and full-time work and does not include hourly wages. In addition, no adjustments are made for the experience and qualification of the employee, and self-employment income is not included in the quarterly statistics.

The QES measures the average hourly and weekly paid earnings of those employed in the construction industry. There are no human capital adjustments for those employed. However, the hourly wage measures allow the quantity of labour supplied by an employee to be controlled. The LCI measures the changes in labour cost for a job with a specified human capital and specified hours of work, that is, it controls for quality and quantity of labour supplied by an employee. The control for human capital and hours worked makes the LCI better suited to measuring the change in wages for individuals working specified number of hours with similar levels of human capital.

Table 2 shows that relative average earnings, as measured in LEED, were virtually unchanged between 2001 and 2009. Average earnings as measured by the QES tended to fall slightly over the period, perhaps reflecting a change in the composition of the construction workforce as an increased number of young people accepted jobs in the industry.

Source: Statistics New Zealand, LEED and QES

Note: Earnings are relative to non-construction (all industries minus construction and not elsewhere classified) in LEED; earnings are relative to all industries rather than all industries minus construction in QES.

Table 3 shows the relative wage inflation using the LCI, which adjusts for such compositional changes, shows that wage inflation in the construction industry was very slightly higher than that for all industries, implying a small increase in relative wages in construction over 2001 to 2009.

Source: Statistics New Zealand, LCI.

Conclusions

In the previous construction boom in the mid-2000s, the construction industry at the national level found many thousands of additional workers each year to meet rapidly rising national demand. These additional workers appear to have been recruited without very large increases in real labour costs relative to other industries.

This relatively smooth national labour market adjustment was in a time when the economy was much stronger than in 2011 and labour was at a premium. There appears to be a large pool of workers not currently working in the construction industry which can move into construction jobs when demand picks up.