World at Work (the compensation professional organization) released recently a compensation budget survey by Compdata Surveys. The big news: the average preliminary pay increase budget is 3.65% for 2007!
Broad averages like this drive me berserk. It is incredibly precise, but downplays the huge variations that affect individual companies and employees.
There is nothing wrong with this average per se. The problem is how it is used. Companies often use average increases like this as a starting point for deciding what pay raises they will give individual employees.
However, like pay, pay increases are determined by the interaction between the local labor market for specific jobs, individual employees' motivations, and a company's business plan. These microeconomic forces dramatically alter the pay increases a company will need to spend, in order to succeed, from what broad macroeconomic averages say.
Companies are free to set pay increases by these broad averages. That is a business management decision. Of course, companies are also free to fail. 🙂
Local variations are what make capitalism fun. In this post, I will look at what data is available, and what forces drive salary increases.
Are you making the most of microeconomic forces to earn what you are worth? Find out in a less than 5 minutes with the PayScale salary survey.
Your taxes at work: CBO predictions for salary increases
If all you care about is broad national averages, then the government is the place to go for predictions. Macroeconomics is the right approach for policy making; the government does a great job of finding the right data to make these broad predictions.
The World at Work (WatW) budget survey percentages basically give the same answers as the Congressional Budget Office’s (CBO) economic forecasts. This is not surprising; the source data for both predictions is much the same: surveys of employers about spending on employees.
The CBO’s increase for 2007 in the “employment cost index” (ECI) is 3.4 percent. The ECI measures the total cash compensation paid by all employers to all employees, divided by the total number of employees in the US. An increase in this index is the same as an increase in an average employee’s wage at an average company.
The WatW survey 2007 pay increase is 0.25 percent higher than the ECI 2007 pay increase. Since these are both predictions about the future, this difference of ~7 percent (0.25/3.4) is close enough for government work.
If average salaries go up 3.6 percent, my raise should be…
What do these broad average pay increase predictions tell a particular company about how much more to pay each of its employees? Not much.
No company employs all workers in all industries. This broad average (even when broken down by industry and location) does not give much guidance to any company, not even huge employers like Wal-Mart.
To see why, let’s look at a related case where broad averages obviously do not apply to specific companies: company sales and the gross domestic product (GDP).
If broad averages applied to specific companies, every company would predict sales increases of 4.3 percent in 2007, because the nominal gross domestic product (GDP) is forecast to grow 4.3 percent. Nominal GDP measures the dollar value of all output from all companies, which is proportional to sales.
I am pretty sure the vice president for sales at PayScale has bigger plans than 4.3 percent growth for this year.
In 2007, companies will see sales growth between -100% (go out of business), and 1000 percent or more (e.g., companies that release their first product). Whether the broad average increase is 2 percent, 3 percent, 4 percent or even 10 percent is not useful for predicting a particular company’s sales growth.
Sales, wages, and GDP: what is the connection?
For a company, sales and wages are obviously connected. The revenue a company earns, by selling what its employees produce, is the source of money for wages. The same holds for national averages like the ECI and GDP.
By looking at the components of nominal GDP growth, we can see both the connection to ECI, and how good the predictions are. The components are (with 2007 increases):
- Inflation: 1.9 percent This is caused by more money chasing the same goods. The Federal Reserve controls the inflation rate by controlling the supply of money.
- Population growth: 0.9 percent Beyond basic demographic trends, the Federal government controls population growth through immigration law and enforcement.
- Productivity gains: 1.5 percent This is how much more people produce in a year, through working longer, harder, and/or smarter. This is not something the government can directly control.
The employment cost index (ECI) leaves out population growth, since it is a per employee measure. The ECI prediction is just the nominal GDP prediction minus the population growth prediction.
This difference between nominal GDP (sales) growth and ECI (wage) growth also makes sense at the company level. If a company were to double sales by hiring twice as many employees, there would be no additional money available for wages per employee.
One caveat: the ECI to nominal GDP increase relationship is only true as long as average pay is a constant fraction of average worker’s output. This average is arithmetic mean, not median: CEOs are employees, so their compensation keeps the average wage up.
Predicting the future: where are the uncertainties?
The government controls inflation, so the government can predict it as well. The government has less control over population growth. However, the core demographics change slowly, and the government has some control over immigration. Population predictions a year into the future are pretty reliable.
The big question mark in future wage and GDP growth is productivity. The government has little control over this. Economists do not even know for sure the causes of the longer term trends in productivity, let alone the year to year variations. Historically, productivity increases have been all over the map.
Productivity is like the weather, what happened yesterday is a good prediction of tomorrow, except when it isn’t. 🙂 Since 1975, productivity gains have averaged ~1.5 percent, so 1.5 percent is a reasonable prediction for 2007. In reality, 0 percent to 3 percent increases are possible.
Employee productivity and company profitability
A company’s employees’ effective productivity can change dramatically from year to year. Productivity varies both in aggregate – all employees at a company work together more or less effectively – and for each employee.
Decreasing productivity shows up as falling profits for a company. Making employees productive – so they produce the most, highest value, stuff that customers then buy – is why CEOs are paid the big bucks.
What we have in the salary budget survey is a precise prediction (<1 percent), which can be wrong by 50 percent, even in the broad average, for example at times of economic downturns.
Worse, the average, because it is a broad average, is not really relevant to companies and employees, because no company or employee is average.
So it goes. I guess a general prediction about the future makes people feel better than none at all. The WatW survey does have value for predicting inflation, but so do the government employment cost index and other measures.
Of course, I am not an economist, so everything in this post is wrong in the details. However, my closest friend from high school is a macroeconomist, so that makes me an expert by proxy.
The PayScale Salary Calculator is a quick and easy way to compare positions. When you want powerful salary data and comparisons customized for your exact position, be sure to build a complete profile by taking PayScale’s full salary survey.
Leave a Reply