Job creation and destruction statistics at the aggregate level reflect the net result of the thousands of businesses in the New Zealand economy changing their specific employment levels from one reference period to the next. This can be due to many reasons, such as the growth or decline of markets, the introduction of new products or processes, competition from domestic or offshore businesses, or adjustments due to changing economic conditions. The analysis of job creation and destruction statistics by dimension (for example, industry) can help explain what causes changes in aggregate labour market movements.
The official job creation and destruction statistics currently published from the Linked Employer-Employee Data (LEED) are produced on a quarterly basis. The calculation of these statistics includes seasonal and temporary variations in employment as job levels are compared at three-month intervals. The variation can be caused by the likes of production cycles, weather patterns, school years, and processing procedures associated with the source data.
Annual Measures of Job Creation and Destruction in New Zealand aims to present experimental annual job creation and destruction statistics from LEED and to examine how these statistics can be applied to complement the existing quarterly measures. The annual statistics are designed to remove seasonal and temporary variations in employment to provide a better indication of permanent job reallocation activity.
Methodology
LEED uses existing administrative data drawn from the taxation system, together with business data from Statistics New Zealand’s Longitudinal Business Frame (LBF). The LEED dataset is created by linking a longitudinal employer series from the LBF to a longitudinal series of employer monthly schedule (EMS) payroll data from Inland Revenue.
Job creation and destruction statistics (commonly referred to as job flow statistics) calculated from LEED are designed to explain the changes in aggregate filled jobs between two specific points in time. The analysis of job flow statistics can help explain what causes changes in aggregate labour market movements. The use of job flow rates (such as job creation, destruction, and allocation rates) can assist in comparing the relative dynamics of employment growth and contraction across different areas of the economy.
Key Results
The annual job creation and destruction statistics presented in this report have been produced using similar definitions and methodology to the existing LEED quarterly measures. However, filled jobs are compared on a March quarter basis one year apart. The annual statistics highlight the strength of the labour market from 2001 to 2008 with positive levels of net employment change, and the downturn in 2009 with a decline in filled jobs. The annualised job creation rate from 2001 to 2009 averaged 14.8 percent, while the job destruction rate averaged 12.3 percent.
The annual statistics at the total level are broadly consistent with those calculated in previous New Zealand studies (Kelly, 2003; Carroll, Hyslop, Mare, Timmins, & Wood, 2002). The statistics can be used to compare New Zealand job reallocation rates with those observed in a recent Organisation for Economic Co-operation and Development (OECD) study (Bassanini & Marianna, 2009). There is evidence to suggest the annual job reallocation rates in New Zealand are comparable with those observed in a number of other OECD countries, such as the United Kingdom and United States.
Annual job creation and destruction rates show a great degree of variation across industry; however, the disparity is not so evident when analysing job flow rates by regional council area. Job flow rates decrease monotonically as the firm size increases; however, the firm size statistics must be analysed with caution given the current method used by Statistics NZ to allocate the firm size dimension (Okolie, 2004).
The annual statistics can be applied to enhance the analysis of the existing quarterly measures by estimating the level of seasonal and temporary variation in employment in the New Zealand economy. The annual method produces significantly less job creation and destruction than by aggregating the quarterly data for the year. Analysis of data from March 2001 to 2009 reveals that on average, over half of jobs created in a year reflect geographic unit employment changes that are offset within the year due to seasonal or temporary factors. A similar percentage of jobs destroyed are also offset within a year. The analysis of industry level data shows the level of seasonal and temporary variation in employment does vary across the economy.