This week, students in the business concentration were gearing up for  Friday’s monthly jobs report with blog posts about what the news was likely to be. They used words like expect, estimate and predict — not always correctly.

It wasn’t the meaning of the words that was off, but rather the constructions that followed them.  For example:

Economists surveyed by Bloomberg estimate personal income to increase 0.4 percent in March.

You don’t estimate something to grow; you estimate that it will grow. Three possible fixes:

Economists surveyed by Bloomberg estimate personal income will increase 0.4 percent in March.

Economists surveyed by Bloomberg expect personal income to increase 0.4 percent in March.

Economists surveyed by Bloomberg expect (that) personal income will increase 0.4 percent in March.

In other words, you have two possible constructions using expect, but only one for estimate.

By the way, that construction would have been fine if the sentence had been written in February. But this was late April. Better: Economists surveyed by Bloomberg estimate that personal income increased 0.4 percent in March. They’re estimating now about something that’s already past.

Here’s a similar construction, but in past tense:

Economists estimated the Case-Shiller City-20 composite to  grow on 1 percent in a the year ending in February 2013.

Economists estimated the Case-Shiller City-20 composite would grow on by 1 percent in the year ending in February 2013. 

Economists expected the Case-Shiller City-20 composite to grow by 1 percent in the year ending in February 2013.

Economists expected the Case-Shiller City-20 composite would grow by 1 percent in the year ending in February 2013.

Since the main verb in the original sentence was in past tense, use would rather than will in their prediction. But in present tense: Economists estimate the Case-Shiller City-20 composite will grow by 1 percent in the year ending in February 2013.

Speaking of predictions: