The good news: Salary increases for U.S. workers in 2014 are forecast to reach their highest levels in six years, according to a new survey released Thursday by the human resources consultancy Aon Hewitt.

The bad news: That increase still won't amount to all that much. And you shouldn't count on it, either.

Forecasted vs. actual salary increases, 2004-2014. DATA: Aon Hewitt, U.S. Salary Increase Survey Forecasted vs. actual salary increases, 2004-2014. DATA: Aon Hewitt, U.S. Salary Increase Survey

Aon Hewitt's 2013-14 U.S. Salary Increase Survey, which includes responses from roughly 1,150 organizations, finds that companies are projecting a 3.0 percent salary increase for 2014, just slightly higher than 2013's 2.9 percent bump and much better than the meager 1.8 percent that was doled out in 2009.

That said, the actual increase hasn't equaled the companies' original forecast since 2007. And while a 3-percent raise may be an improvement over recent years, it's still not up to pre-recession levels, when 3.6- and 3.7-percent increases were common.

The big reason across-the-board raises are still a little weak is, of course, leaders' apprehension about the economy. But it's also due to more and more payroll dollars going toward performance-based awards such as bonuses and profit sharing. Twenty years ago, companies were sending 5.9 percent of their payroll budget toward such "variable pay." Today, that has doubled, Aon Hewitt found, with companies allotting 12 percent of their budgets to it.

Lower salary raises for everyone and bigger bonuses for the top performers may make rational sense on paper. Give more of the pool to the top performers, the theory goes, and they'll return the favor with even greater productivity. But in practice, it's not clear such a system is always effective. Only 22 percent of the survey's respondents said they give most of their "variable pay" budget to high performers, some to average performers, and none to the worst performers.

Moreover, the differences in merit-based salary raises for top and middling performers can be so negligible one has to question their potency. People who got the best possible performance rating, Aon Hewitt's survey found, saw a raise of 4.7 percent on average, while the majority of employees who simply "met expectations" saw a 2.6 percent raise. Do the math on a $75,000 annual salary, and the difference between the bump the best people see and the one the average folks get is just $1,500 a year, or an additional $60 per biweekly paycheck (and that's before taxes). A family of four can barely manage a dinner out on that.

Here's the question leaders should be asking about this survey: Is such a meager boost for the best worth the hit it has on the rest?

Merit based salary increases for different performance ratings DATA: Aon Hewitt, U.S. Salary Increase Survey

Jena McGregor is a columnist for On Leadership.

Read also:

Why slackers still get bonuses

What's better than a bonus? The award you give away

Like On Leadership? Follow us on Facebook and Twitter.