Technical Note for Sample Variance of New Jersey and Metropolitan Statistical Area (MSA) Estimates of Monthly Nonfarm Payroll Employment 

Monthly estimates of nonfarm employment for the state of New Jersey and its Metropolitan Statistical Areas (MSA) come from the Current Employment Statistics (CES) survey of employers. The CES survey is conducted by the New Jersey Department of Labor and Workforce Development cooperation with the US Bureau of Labor Statistics. The CES survey, like other sample surveys, is subject to two types of error—sampling and nonsampling.  The magnitude of sampling error, or variance, is directly related to the size of the sample and the percentage of universe coverage achieved by the sample. Measurements of error associated with sample estimates are provided in table 1.

CES employment estimates for New Jersey and its MSAs are produced using two methods.  The majority of New Jersey statewide and area estimates are produced using direct sample-based estimation.  However, a few published estimates that do not have a large enough sample to support estimation using only sample responses have been calculated using a small domain model. Direct sample-based estimation allows for the computation of sample error variance while small domain model estimation does not. Sample errors have been computed for total private sector employment and for all supersectors statewide. 

Variance of the CES estimates.  The sampling errors provided in this technical note have been estimated using past employment information from the Quarterly Census of Employment and Wage program.  The Covered Employment and Wages or QCEW program provides a near universe count of employment derived from Unemployment Insurance tax records.  The Unemployment Insurance data serves as both the sample frame and annual benchmark source for CES estimates. For an explanation of benchmarking, see the article "Major Charges in Nonfarm Wage and Salary Employment”, pages 9-16 in the March 2003 issue of Economic Indicators.  The sampling errors are based on average response rates and employment reporting patterns from past Unemployment Insurance reports.  Under normal circumstances, where the average response rates are met and there are no reporting anomalies, the sampling errors represent a good indication of the magnitude of error in the estimates from sampling only.  They do not represent components of error in the estimates that are due to reasons other than sampling; nonsampling error is captured in the benchmark revision.

Appropriate uses of sampling variances in CES.  Variance statistics are useful for comparison purposes, but they do have some limitations.  Variances reflect the error component of the estimates that is due to surveying only a subset of the population, rather than conducting a complete count of the entire population.   However, they do not reflect nonsampling error, such as response errors, and bias due to nonresponse.  The overall performance of the all-employee estimates is still measured in terms of the benchmark revisions.  However, variances are very useful in determining when changes such as over-the-month, quarter or year are significant at some level of confidence.

It should also be noted that extremely small estimates of 2,000 employees or less are potentially subject to large revisions that are caused by occurrences such as the relocation of one or two businesses, or a change in the activities of one or two businesses.  These are non-economic classification changes that relate to the activity or location of businesses and will be present for sample based estimates.

Illustration of the use of table 1.  Table 1 provides a reference for standard errors of New Jersey and area employment estimates.   The standard errors of differences between estimates in 2 non-overlapping industries are calculated as

The errors are presented as standard errors.

Table 1 provides a reference for the standard errors of 1-, 3-, and 12-month changes in employment. The errors are presented as standard errors of the changes.

Suppose that the over-the-month change in manufacturing employment from January to February in a given state is 1,500.  The table indicates that the standard error for a 1-month change in employment for manufacturing is 1,200.   The interval estimate of the over-the-month change in employment that will include the true over-the-month change with 90-percent confidence is calculated:

1,500 +/- 1.645(1,200)   =   1,500 +/- 1,974 = [-474, 3,474]

The true value of the over-the-month change is in the interval -474 to 3,474.   Because this interval includes 0 (no change), the change of 1,500 shown is not significant at the 90-percent confidence level.   Alternatively, the estimated change of 1,500 does not exceed 1,974 (1.645 x 1,200); therefore, one could conclude from these data that the change is not significant at the 90-percent confidence level.

Table 1.  Standard errors associated with 1-, 3-, and 12-month changes in employment for NAICS supersectors

State/Area	Industry	Standard Error on 1 month change	Standard Error on 3 month change	Standard Error on 12 month change
New Jersey State Estimates	Total Nonfarm	8,061	14,330	24,134
	Total Private	7,322	13,035	21,856
	Goods Producing	2,401	4,339	6,906
	Private Service-Providing	6,791	12,191	20,219
	Mining and Logging	16	32	74
	Construction	2,040	3,612	5,565
	Manufacturing	1,253	2,062	4,087
	Durable Goods	945	1,434	3,386
	Non-Durable Goods	809	1,446	2,363
	Trade,Transportation, and Utilities	3,551	6,831	11,155
	Wholesale Trade	1,029	2,081	3,311
	Retail Trade	2,440	4,119	6,158
	Transportation, Warehousing and Utilities	1,499	3,157	8,636
	Information	781	1,317	2,420
	Financial Activities	1,517	2,863	5,952
	Professional and Business Services	2,981	5,768	9,253
	Educational and Health Services	2,069	4,741	7,631
	Leisure and Hospitality	2,438	4,580	7,337
	Other Services	1,959	3,291	4,119

a- Less than 80 percent of employment is estimated by direct sample-based estimation.

For more information on variances including data for other states, see http://www.bls.gov/sae/790stderr.htm.