Thought leadership

How to Select and Use Labor Market Data

How to Select and Use Labor Market Data

Two of the core strategies in managing an organization’s human capital base are to pay both competitively and cost-effectively.  Sounds simple, but unless you know how to select and properly use local labor market data for making pay decisions, you may be underpaying some and overpaying other employees.

The purpose of this paper is to help employers pay people right.  This paper is organized into two major sections.  The first addresses many of the frequently asked questions about labor market data and the second deals with how to use those data.

What is the labor market?

The labor market is simply a measure of the “going rates” of pay for different types of jobs based on the supply of and demand for employees with the skill sets required to hold those jobs.  For any given job, the market rate will vary depending on location, company size, and/or industry, to varying degrees.  Executive jobs are particularly sensitive to company size.  Clerical, administrative, technical support and some professional and middle management jobs tend to be more sensitive to location.  For example, an entry level accountant may be paid 8-12% more in Oakland than in Stockton, CA — a short 90 miles away.

Why are labor market data important?

As an employer, it is very important to pay competitive wages and salaries to attract and retain employees.  The key is to define what “competitive” means in your organization.  What is competitive is usually a function of a number of complex, and sometimes intangible factors, including how relatively easy or difficult it is to fill open positions, what the organization’s financial situation is (i.e., how much the organization is able to pay) and how well the compensation function is managed.

Labor market data are also important because they help with recruiting new staff and estimating compensation costs for an annual budget plan.

What is meant by “managing” compensation?

For most employers, employee compensation costs are the single largest expense item.  It’s not unusual for payroll to account for 50% or more of a firm’s operating budget.  Therefore, it is essential that wages and salaries be neither too high nor too low.  Based on our experience with more than 1,000 client projects, most companies both overpay and underpay.

The solution to this problem is to actively manage pay decisions.  Without reliable labor market data on most of the organization’s jobs, employers may find themselves not only mis-estimating salaries, but in the position of being held hostage to employees’ demands, anecdotal information, and managers’ supposed knowledge of the market.

What is labor market inflation?

As noted before, the labor market is defined as the supply of, and demand for, people . . . and is indicative of what employers need to pay in order to be competitive.  Labor market inflation is the rate at which salaries and wages increase over time.  Inflation rates vary by job — not all jobs inflate at the same rate over time.

Why not use cost of living increases to determine pay adjustments?

Cost of living indicators, like the Consumer Price Index (CPI), measure increases in prices of a market basket of consumer goods and services.  They do not necessarily reflect actual market-rate wage and salary increases.  However, this practice continues in many organizations because it’s easy and sounds logical on the surface, especially if it’s historically been handled this way.

Why do we focus on location specificity?

Many compensation analysts have commented that it’s getting more and more difficult to find reliable pay data that relate specifically to different locations within a major metropolitan area.  Some people have perceptions that salaries are higher, for example, in Silicon Valley than in Marin County, or in Los Angeles compared to San Diego.  Without specific job-by-job data by location, one cannot know.

We have, however, observed that fewer employers are willing to take the time to submit their cash compensation and benefits information for inclusion in national or regional salary surveys.  As a result, significant erosion in both sample sizes and data specificity have occurred.

That’s why in 2003 we started the Greater San Francisco Bay Area Compensation Survey and in 2011 introduced the Gallagher California Total Compensation Survey, which captures job data by zip code.

How many surveys should be used?

Compensation professionals strongly advocate using multiple sources of survey data to create a “labor market composite” for location-specific jobs.  Every data source has inherent limitations, so using multiple sources provides you with a refined estimate of the true market range for each job.

Our general rule of thumb is to use at least three to five different surveys per job for optimizing pay decisions.

Where can data on industry-specific jobs be found?

Many compensation surveys include jobs that are typically found across all kinds of organizations — plus jobs that are found only in particular industries or types of organizations, such as fundraisers in non-profits or assembly workers in manufacturing companies.
Organizations should supplement any generic surveys they have by using additional published surveys that provide data on jobs that are unique to their industry and/or type of endeavor.
It is also recommended that, in the absence of adequate published data, employers consider engaging the services of a third party compensation consulting firm to conduct a custom survey designed to fill their unique data needs.

What’s the importance of current pay data?

One strategic part of your competitive advantage as an employer relates to pay.  It is imperative that you be able to access reliable sources of pay data that are specific to your location, jobs, and organizational size.  In today’s economy, using current data to make informed pay decisions is more important than ever.  In addition, data should always be aged using reliable estimates of labor market inflation.

Are there problems with today’s pay data?

Over the last several years it has been apparent that the reported ranges of actual salaries have become wider, an indication that: (1) organizations are setting salaries without the benefit of explicit labor market indicators, and/or (2) job matching in connection with survey participation has become significantly less disciplined than it was in the past.

In addition, managers are now faced with increasing challenges from employees due to their use of Internet sources for pay data.  Most experienced compensation managers understand the shortcomings associated with using Web-based employee-generated information.  However, others seem to be fascinated with the availability of free data.

Unfortunately, if you rely only on free data, you might be increasingly unable to respond in the best interests of your organization.  It’s important to use statistically valid information submitted by employers, not employees, on job-specific wages, salaries and other cash compensation elements.

What types of surveys are available?

Published Surveys are compilations of base pay and bonus data, reported by participating employers.  Generally used for benchmark positions, many surveys are specific to industry, location, or size of organization.

Custom Surveys are specialized surveys that focus on a select set of organizations within a particular industry and/or location.  They usually obtain pay data on particular industry-specific or rarely surveyed jobs.

Internet-based Surveys are generally considered unreliable by compensation professionals because these data are reported by employees, not employers.  This is especially true for free Web surveys.

Choosing and Using Labor Market Data

Listed below are the steps involved in the process of acquiring and using labor market data when developing a broad-based structured base pay plan. 

1. Assess the data needs of your organization.

Identify which organizations are competitors for people, by type.  For example, if you are in the insurance business, you are competing with other insurance organizations for claims examiners, but you’re also competing with all other employers in your geographic area for administrative assistants.

This means that it’s important to assess your data needs in terms of which jobs are “generic” or core, i.e., found in all kinds of firms, and which ones are industry-specific, cognizant of the particular employer segments within which you compete for people.

2. Identify and acquire survey data sources.

As noted in the first section of this white paper, use multiple surveys to effectively assess competitive pay levels applicable to your particular situation.  Rarely does one survey meet all the data needs for an organization.  Surveys that focus on a particular industry often also include generic jobs.  But sometimes data on these jobs may be unrealistically high or low.

For example, when technology companies were paying extraordinarily high salaries for software development engineers from 1999 to 2001, salaries for clerical jobs in that sector rose as well, even though the total supply and demand picture for the metro area was quite different.  In April 2011, Google announced that all employees, including those in clerical jobs, would get a 10% pay increase.  That doesn’t mean all clericals in the San Francisco Bay Area or Silicon Valley should be paid 10% more.  These aberrations are just distractions from the analysis of the real markets.

Most organizations should utilize both generic (core), location-specific and industry-specific data sources.  These sources may be purchased from trade and industry associations, local business organizations, membership organizations or compensation consulting firms.  For hard-to-find jobs, it’s sometimes necessary to sponsor and/or participate in a specially-focused custom survey.

3. Select “benchmark” jobs.

Benchmark jobs are the ones that are populated by the most employees (multi-incumbent), and those for which surveys are readily available.  It is not necessary to find labor market data on every job in your organization if you apply a job evaluation process to determine internal relative job values.  The resulting internal job levels will enable you to “slot” the unique jobs — those for which data are not available — in between or equivalent to jobs for which you do have valid market data.

4. Match jobs to survey descriptions.

Always look for survey data that include job descriptions as matching criteria.  Titles can be very misleading, as they are often names for very different sets of job duties from one organization to another.  It is also critical that your job descriptions are up-to-date when matching.

Because organizations are living, breathing entities, it is not unusual for 30% – 40% of job descriptions to be out-of-date.

Job Families

In cases where surveys report information on different levels in job families, required experience and education for the job can also be considered as differentiators, but job content should always be the primary matching priority.

Hybrid Jobs

These jobs, which are generally unique to the organization, are those that have responsibilities in different functional areas and usually cannot be found in surveys.  In these situations, match the individual components of hybrid jobs to survey job descriptions to get a composite sense of the range of pay for the job.

5. Don’t forget to age data.

Plan ahead.  Estimate how overall labor market inflation projections will affect salaries at the midway point of the period for which you plan to use the market data.

For example, if your salary structure is intended to be in effect for a full calendar year, the data can be aged to July 1st of that year.  That way, your salaries will slightly lead the market for the first half of the year and slightly lag it for the remainder of the year, resulting in reasonably competitive pay levels throughout the period.  However, it should be recognized that aged data are estimates, not scientific facts.

Aging is also a way to compare data from different sources, with different effective dates, to one common point in time.

Keep in mind that labor market inflation reflects how much actual base pay levels are increasing over time.  This is a function of labor market supply and demand.  This is not, as mentioned before, the same kind of inflation as that which is measured by the Consumer Price Index (CPI).

Also, most employers are generally more concerned about paying competitively and cost-effectively than they are with keeping their employees’ purchasing power whole.  Therefore, using measures of local labor market inflation usually makes more sense than applying the CPI, or Consumer Price Index, to determine pay adjustments. 

6. Recognize that data for any given job will comprise a range.

The analysis of market data from multiple sources will result in a range of actual pay levels.  Because data collection is not an exact science, it is to be expected that data from different sources will reflect different segments of the market.

Sometimes, the pay ranges will be quite broad, especially for highly paid positions, and narrower for lower priced jobs.  It’s important to identify where in the market range the organization wishes and/or needs to pay.  Internal job analysis will help in this process.


We hope the information discussed in this paper will help you make better pay decisions.  The six points below are some of the key things to remember when analyzing and using labor market data:

  • It’s important to keep your job descriptions current.
  • Use multiple survey sources for each job.
  • Use specific matching criteria: location, revenue and industry.
  • Use multiple data “cuts” from individual surveys when they are available.
  • Surveys have different effective dates, so age data to a common date.
  • Use internal relative job values for jobs for which you have no data. 
Shari Dunn

Upon receiving her B.A. degree in psychology from the University of California at Berkeley, Ms. Dunn headed to New York City to take a position in the human resources department of the Marine Midland Bank, and later worked at Nabisco in operations research. Returning to the San Francisco Bay Area, she became a Research Associate for McKinsey & Co. Next, she moved to Deloitte (then Touche Ross) as a consultant. It was at these two ...

Read Full Bio