Read the PDF Version (PDF)

We all know that salary surveys are a key part of the process for making salary decisions. But do we really know how to determine if salary survey data is reliable? With the fairly recent access to salary data on the internet, there are even more reasons to question reliability of salary information which may be end up on your desk from coworkers.

You may have discovered that different salary surveys often provide conflicting or widely varying data. This is one of the main reasons a market based compensation system (one that is developed strictly from salary surveys) is difficult to maintain or administer. The conflict comes from three factors associated with collecting the data: the size of the sample, the closeness of match in the jobs being surveyed and the selection of the organizations responding to the survey

Critical Factors

The size of the sample is a critical factor in the reliability of any salary survey. Small sample sizes are simply not reliable because they do not provide a large enough cross-section of the population to be representative of that population. They also tend to have radical changes from year to year unless the exact matched sample is selected each time. Best practice is to only use surveys that have a large enough sample to make the data a reliable predictor of the overall population.

Also, equally as critical, is the closeness of the match of the job’s being surveyed. The same job title can be very different in job content, level of responsibility, and complexity between different industries or organization sizes. Most surveys attempt to limit the variance in job content but do not generally deal effectively with differences in level of responsibility and complexity. If you have completed a salary survey you know how difficult they can be to complete. Many people simply resort to reporting salaries based on the job title without matching the job content, which can be misleading and cause significant variability in the results.

Additionally, who responds to a survey can also influence the results. The size, industry and location of the organization can dramatically influence the data being used for comparison. It is critical that data for similar size organizations are used when developing the pay ranges for the middle managers and executives. It is also important to use national data in the collection process so you have a consistent reference point to compare to when applying adjustments to reflect the local or regional market levels.

Ageing Data – Is it Reliable?

Once you have established the salary survey data and resources for your organization, it is important to be consistent and use the same surveys from one year to the next. Depending on when compensation data was collected, the data may need to be “aged” to account for changes in the wage market since its original collection. Ageing data is a practice that is complex and requires considerable expertise and thought going into the process.

A simple way to do this is to take the annual rate at which salaries are moving for this job and prorate it, salary increases overall this year are around 3.5% but this may vary by job title. Another method, which is more likely to predict actual salary movement, is to access “inflation index” information such as the Consumer Price or the Producer Price Indices. Keep in mind that the Producer Price Index (PPI) measures price change from the point of the seller whereas the Consumer Price Index (CPI).measures price change as viewed by the purchaser.

In any case, ageing data for more than one year is not recommended as other factors affecting salaries for jobs, such as supply and demand, would not be reflected when simply increasing salaries or ranges by a flat percentage. To ensure you have valid, reliable data, best practice is to purchase the survey data from one year to the next rather than relying on “ageing” to update salaries and ranges.

Benchmark Jobs

Benchmark jobs are jobs that tend to be consistent within the industry and across multiple surveys. It is important to identify benchmark jobs within your organization and establishing salary ranges for those jobs. The salary rate established for each benchmark job is the best estimate of the market rate for these jobs.

It is advisable to collect salary data for each of your benchmark jobs from a variety of national survey sources to establish a weighted average national salary for each benchmark job. The weighted average national salary data should then be adjusted to reflect geographic wage differentials for the area in which you recruit your employees, whether it is local, regional or national. Geographic differentials are developed for hundreds of locations throughout the United States and Canada by compiling and comparing salary survey data from hundreds of local area surveys conducted by government, industry, trade associations and private business. Geographic wage differentials are used by all major U.S. corporations to establish both wage and cost-of-living standards for their branch locations.

Salary Ranges & Compa-Ratios

Salary ranges should be established that are a fair reflection of the market and industry in which you compete. The range structure should be adjusted each year to reflect the actual market movement that is identified by analyzing the benchmark market survey data.

Compensation is administered through management of compa-ratios. The compa-ratio is a mathematical calculation that identifies the relationship between actual pay and market rate for a person, department or the total organization. For example, a 90% compa-ratio for an individual says that the individual is being paid at 90% of the prevailing market rate for his/her job. The compa-ratio is determined by dividing actual pay by the rate established as the mid-point for the grade.

The mid-point of each pay grade reflects the median pay of people in the same or comparable jobs within the relevant labor market/industry. Approximately one-half the people in comparable jobs are paid above the mid-point and one-half are paid below. The minimum of the range typically is 20% to 30% below the mid-point and the maximum is typically 20% to 30% above the mid-point.

Compa-ratios between 80% - 90% are considered entry rates. New employees are generally hired at these rates. Compa-ratios between 90% - 97% would be considered appropriate for employees who are not yet fully trained and qualified for the position. Compa-ratios between 97% - 103% would be considered appropriate for employees who are fully qualified for the position and who, over time, consistently perform at an acceptable level. Compa-ratios above 103% would be appropriate for employees who are fully qualified, and over time, consistently perform above acceptable levels. As a general rule, most employees should fall within the 97% - 103% level within 2 to 5 years of experience.

As you can see, obtaining, evaluating and administering market data can be a complicated process. It is much more than simply completing a search on the internet and plugging those numbers into your salary structure. If the end result is to create a fair and equitable system, many other considerations must be included and the refined process may take some time to establish. But in the end, it is an important tool in recruiting and retaining employees so the dollars spent will certainly pay off dividends in the end.