Employment Level

The Employment Level is Switzerland's Employment Barometer, the quarterly count of jobs from BESTA, the Job Statistics survey of businesses. The Federal Statistical Office (FSO), Switzerland's national statistics office, runs it quarterly since 1925, in headcounts since the third quarter of 1991. It covers jobs subject to OASI payroll contributions, the Swiss social-security system, in industry, construction and services on Swiss territory. It excludes unemployment, employed people, farm jobs, private households and short-time work, and one person with two jobs counts twice.

Responses cover jobs in the final month of each quarter, and results appear about 8 weeks later at 08:30 Swiss time. The first quarter arrives in late May, the second in late August, the third in late November and the fourth in late February. SNB, the central bank, and SECO, the labour office, see data 5 working days early. Funding cantons and cities see regional data two days early and agencies 15 minutes early.

The sample is renewed every two years with the prior two years revised. The adjusted series is rebuilt each quarter, and outlook above 1.0 reads as positive.

Moves come from industry, services, hours worked, unfilled posts and firms' hiring signals.

• Secondary sector: captures industry and construction jobs, classified by NOGA, the Swiss activity classification.
• Tertiary sector: captures services jobs such as trade, transport, finance, education and health.
• Full-time equivalents: capture work volume converted into full-time jobs, so hours shifts show beyond headcount.
• Vacancies and vacancy share: capture unfilled posts at quarter end and their share of filled plus vacant jobs.
• Employment-outlook indicator: captures firms' plans to maintain, increase or cut headcount next quarter on a 0.50 to 1.50 scale, where 1.0 means maintain.
• Recruitment difficulties: capture how often sought staff were found with difficulty or not at all, a friction gauge.

In the third quarter of 2019 jobs rose 1.3% from a year earlier, up 67000 to 5137000, the highest level since the series began. Industry and construction added 1.2% and services 1.4%, while work volume rose 1.1% and adjusted jobs rose 0.3% on the quarter. Yet the hiring outlook fell to 1.03 and recruitment difficulties eased to 31.8%, the first decline since the fourth quarter of 2016.

Vacancies still grew 9.6% to 79000. The lesson is that a record job count can coincide with softer forward signals, so headline level and expectations can diverge.

1. Sampling: a random sample grouped by industry is drawn from the business register, then all sites of chosen firms enter.
2. Observation: firms report jobs in the final month of the quarter in 4 working-time bands, including weeks under 6 hours. They also report work volume, end-quarter vacancies, hiring plans and hiring frictions.
3. Non-response treatment: Missing giants (very large firms) are filled from prior data with sector trends, and other weights are raised to cover non-respondents. Weights are then aligned to register totals and extreme values are dampened.
4. Extrapolation (scaling to population totals): weighted sums scale the sample to population totals, with total jobs as the sum of the 4 bands. Work volume is estimated directly from sample weights.
5. Derived rates: Year-on-year change divides the 4-quarter job difference by jobs 4 quarters earlier, while quarterly change uses X12-ARIMA, a seasonal filter rebuilt each quarter. Vacancy share divides vacancies by filled plus vacant jobs, and the outlook scale runs 0.50 for cut, 1.0 for maintain and 1.50 for add.
6. Illustrative example: (5137000 minus 5070000) divided by 5070000 times 100 is about 1.3%, or 67000 jobs. Also, 79000 vacancies divided by 5216000 jobs plus vacancies times 100 is about 1.5%, and 1.03 sits just above 1.0, or maintain.