Salary Raises: Understanding Cost of Living vs. Merit Pay Increases

I have been looking at pay raises lately, and one thing I have been struggling with is how HR managers talk about raises.

For example, a recent question on an HR discussion board was whether companies should give cost of living adjustments (COLAs) or "merit" increases.

In this post, I'll cover what HR means by COLAs and merit raises, and then discuss why this is not really the right way to talk about pay and raises.

Are you due for a raise? Use PayScale's detailed salary calculator to figure out based on your unique skills and abilities what you are worth.

What are COLAs and Merit Raises?

Cost of Living Adjustments mean exactly that: what you are paid goes up based on the cost of the goods and services a typical person buys. COLAs make complete sense for entitlements like social security.

The idea of social security is that the government is giving you money to buy a basket of goods to survive. If that basket of good costs more due to inflation (an increase in the cost of living), then you need more money.

For the youngsters among the readers who don't remember double digit annual inflation, before everything was indexed to inflation in the early 80's, retirees would regularly see their buying power go down 10% or more each year. It took an act of Congress to raise the payments they would receive.

Merit Increases are an internally focused raise philosophy. Managers rate their employees (or employees rate each other in a "360" evaluation philosophy), usually based on performance over the last year.

Top performers get a larger raise, while the bottom performers get no raise. The raises are supposed to be motivational. Top performers will work hard for a larger raise while no raise is a warning to low performing employees that they better shape up.

The Forces that Drive Pay

COLAs and merit increases confuse me because they are largely disconnected from the external market and internal value propositions that drive pay.

There are only two forces that drive pay:

  • Market Price: what the employee is worth in the market
  • Internal Value: what the employee is worth to the company that pays them

If all non-cash factors are the same โ€“ quality of management, career advancement potential, benefits, etc. โ€“ companies are free to pay above the market price based on internal value. They cannot pay much below the market price for long and keep workers, whether the workers are average or outstanding.

Raises, no matter what they are called, are about recognizing changes in market price and/or internal value of a worker.

Changes that Justify Raises

There are three distinct types of changes that can affect pay. In the follow sections, we will look at each in turn.

Case 1: An employee's characteristics change in a way that affects his/her market price

For example, the typical software developer with 3 years of experience is worth more than one with 2 years of experience. This is a fact in the market โ€“ in the Seattle area, 3 vs. 2 years of experience is good for about 5% more pay.

Any employer that does not recognize this as an annual "merit" increase or promotion with increased pay is at risk of falling below the market price for this growing employee.

Note that this is not about being a "top" performer. The average developer with 2 years of experience should earn 5% more pay the next year.

Case 2: The market price for the employee's skills has changed

For example, the demand for occupational therapists remains hot nationwide, even in this recession. If a company has an OT who is fully competent in the role (e.g., 10 years of experience or more, so no increase in competence or productivity from last year), the company may need to pay 5% more to keep her/him simply because of the market demand pressure on wages.

This brings up why I don't like "cost of living" adjustments: just because the price of oil goes up does not mean wages for a particular job should go up. The correct factor to consider is market price changes for workers in a job.

Market forces can cause the pay of workers to go down.

For example, PayScale data indicate the typical college graduate-level employee has seen total cash comp. go down about 2% in the last year, even though the consumer price index (CPI: a standard measure of changes in cost of living) in the same period went up 2.3%.

This drop in pay reflects the weak employment market in the US across many jobs. A COLA based on the CPI would be overpaying by about 4% on average.

Have employers really cut base salary 2% in the last year? Most of this 2% in pay cuts has come through reduced bonuses, laid-off employees accepting new jobs at lower wages, and reduced profit sharing. However, it is not hard to find employers who actually cut base salary for at least some employees. For example, the University of California cut faculty pay 4 to 11% in the last year.

Typical real wage cuts (reductions in buying power) of 4% a year or more during a recession are not uncommon. World at Work salary budget data shows that, during the last major recession of the early 80's, wage increases generally were at least 4% below CPI each year. When inflation is 9%, a 5% pay increase is a real wage cut.

Case 3: The value of the employee's work to the company has changed

For example, a project manager brings in a particularly difficult project on time and under budget, earning the company substantially larger profits.

If this performance represents an increase in competence that other employers will recognize, then this is just another example of case 1): market price has increased. A company needs to increase pay by at least what the market would pay for this increase in competence.

If the market value of the employee's productivity is not so clear, companies have two options. If they don't think this increase in effectiveness will happen every year, paying a one-time bonus to recognize a job well done may make sense.

If this work is indicative of the employee actually representing the cream of his/her comparison group, then this may be a case for targeting a higher percentile pay in the market.

What is the Market Price for a Job/Employee?

Despite my love of market data, a particular worker's pay can vary a lot and still be "at the market". For example, even a well-described professional position or employee will likely have a difference between the 25th and 75th percentile (bottom 25% and top 25%) of the market rate of 30% or more.

Employees care about 30% differences in pay: paying $50,000/year or $65,000 matters :-)

Companies often target the 50th percentile (median) market pay for a job. However, if all measurable factors were taken into account in pricing the employee or job in the market, there is no particular reason the median is a better target than the 75th percentile or 25th percentile.

After all, 1 in 4 of these "identical" workers make above the 75th percentile, so that level of pay is not uncommon.

The variation in pay in the market between the top and bottom percentiles often represents hard to describe (and price) productivity differences between individual employees and across companies.

What market percentile a company should target is a business plan decision.

For example, Netflix's approach is to pay near the 90th percentile (top 10%) of market, or the employee's internval value to Netflix, which ever is higher, for nearly every key employee. Anyone who leaves Netflix will likely take a pay cut to do the same job elsewhere. Netflix figures higher worker productivity of these top performers will make up for the greater cost.

Other employers target a low market percentile, e.g., the 25th percentile or below. (Wal-Mart comes to mind). These employers want to pay the least they can to control the cost of labor. These firms usually have efficient mechanisms for hiring workers and bringing them up to speed quickly, because they generally have high turn-over.

Knowing your company's target market price philosophy goes a long way to helping you understand how your company thinks about workers and worker productivity.

COLAs or Merit Increases?

So which should a company pay, COLAs or merit increases? The answer is neither.

Changes in the market price for a company's workers, measured by compensation surveys like PayScale's, are what should drive changes in workers' pay.

COLAs have basically nothing to do with the market value of a worker's skills and abilities. During times of recession, annual wage increases well below annual inflation are common. For hot jobs in high demand, raises well above inflation are needed to keep workers.

Similarly, merit increases don't work when they are not tied to the market. If a company's annual review determines a worker is a top performer, but that worker is already paid at the top of the market (at a high percentile) for his/her skills, then no raise is a reasonable answer.

Conversely, for many junior employees, even an average worker is growing quickly in competence, so needs regular, substantial, pay raises to keep up with her/his increasing market value.

No matter whether your are a junior employee, or aiming for retirement, are you earning what you are worth? For powerful salary data and comparisons customized for your exact position or job offer, build a complete profile by completing PayScale's Full Salary Survey.

Cheers,

Al Lee

4 Comments

  1. 4 Steve 10 Oct
    Every corporation I've worked at has based raises on current salary, not on what the market was doing. That has two implication: 1) The only way to get a big raise and/or catch up to the market is generally to get a new job at a new employer. 2) The hours and hours of extra work the other commenters put in pale in comparison to a few minutes of negotiating the starting salary when getting that new job.
  2. 3 Bill 10 Sep
    When employers do away with HR departments, and discontinue all yearly evaluations or reviews, it allows them to force employees to come to them and plead their case after "however many years" go by. The problem with this whole article AND with using Payscale at all, is because it has no method of figuring in your value, when you cover many job descriptions. If you are an administrative assistant, but have a knack for I.T., and truly support the company, its network, its employees performing work that I.T. professionals are paid a salary to perform, you cannot get any kind of idea as to IF or SHOULD this admin asst be paid a higher wage for the additional job functions. Payscale.com has no method of computing this. I work for a sole proprietor that has made up his own rules, who exploits the talents of employees outside of their unwritten job descriptions, so that he does not have to pay for it. Crooked and lacking integrity. I think Payscale.come needs to align itself with the employees rather than the tyranny of poor- management.
  3. 2 jon 03 Sep
    I had the same increase in raise as said above by Dan. My raise was 2.43, how that got that number I dont know? But why give me a petty raise for working past expectations when COLA is 2%. Working harder does not justify to a .43% raise of hell I had to deal with last year. Most likely they figure time vs money. If you're salary or hourly. Still 2% is barely COLA if you are an exceptional worker.
  4. 1 Dan 11 Jul

    If I received a "merit" increase of 2.5% for "exceeding expectations", would I be correct to factor out the inflation rate since my last performance review to come up with the more realistic "net" increase? (i.e. if inflation rose 2% year-over-year since my last review, would I be correct in saying that I really only got a .5% raise for "exceeding expectations"). Yes, they could have easily given me nothing, but I kicked my ass to "exceed expectations" this past year and don't see much of a reward for it, especially after adjusting it for inflation. I would have rather had a way better work-life balance over the last year and simply "met expectations" and received nothing in terms of a raise.

    What do you think? If a company does not give COLAs, am I correct to account for inflation myself after getting a raise based on a regular merit review?

Comment




  1. Please prove to us that you're not a robot: