Over the weekend, Kevin Kelleher of the GigOm blog network argued that Google (NASDAQ:GOOG) should start paying a dividend to investors:

Five years ago this April, Google filed to list its stock publicly. The founders let potential investors know it wouldn't play by some of Wall Street's rules, including paying them a cash dividend -- which, the prospectus boasted, Google had never done. And as of today, it still never has. But maybe it should -- at least until things turn around for Google.

At least until things turn around? Bad idea. Dividends aren't a stop-gap, payments that can be turned on and off like a faucet. Look at history. Once issued, dividends become expected payments -- stop paying, or reduce the payout, and investors will revolt.

Whither dividends?
And while it could be a coincidence, last year was one of the worst markets on record and the worst year for dividends in five decades. Cutters such as Fifth Third Bancorp (NASDAQ:FITB) and Freeport McMoRan Copper & Gold (NYSE:FCX) badly lagged the market's 40% loss during 2008.

I don't think that Kelleher is wrong to ask whether Google should pay a dividend. Plenty of Google's tech-tastic peers pay one, including Intel (NASDAQ:INTC), Microsoft (NASDAQ:MSFT), and IBM (NYSE:IBM). It's perfectly fair to consider whether Google should join the party.

My issue is with how he poses the question. Kelleher assumes, independent of further financial analysis, that because Google has the cash and the heft to pay a dividend, it should. "The biggest obstacle to Google paying a dividend is more of an emotional one -- corporate pride. Dividends are what mature, aging companies pay to keep angry investors at bay," he writes.

Corporate pride? Really?

I doubt it. Google executives have pride, to be sure; just look at their aggressive pursuit of Microsoft. But citing pride as a barrier to a dividend ignores some important fiscal truths about the Big G. Top of the list is execution; management has a history of investing well. Take its most recent quarter. Even as Google shut down development on six nascent projects, cost cuts yielded a better-than-expected earnings result.

And for all its failures, when Google wins, it wins huge. Google Maps and Google Docs, for example -- both of these services were, at one time, projects within the Big G's Labs.

Think of Google as you would an expert poker player. No doubt it'll lose some big hands, but by investing heaviest when the odds favor success, it produces better-than-average returns on equity and invested capital:

Metric

2008

2007

2006

2005

Return on invested capital

16.6%

16%

16.8%

21.3%

Return on equity

16.3%

21%

23.3%

23.7%

Data current as of Feb. 22, 2009.

Yes, those numbers are declining, but they still dwarf the stats put forth by Yahoo! (NASDAQ:YHOO), its closest peer.

Bottom line: Google is a value creator. Why take retained earnings away from a company that's proven it can produce returns for the sake of a short-term dividend? It doesn't make sense.

On the other hand ...
Nevertheless, Kelleher is right that Google shareholders have suffered over the past year. A dividend could offer relief to battered portfolios.

You could also argue that, given a planned $460 million stock options repricing plan -- a bold wealth transfer from shareholders to employees -- now is the perfect time for a one-time dividend. Make it $460 million precisely, $1.46 for all 315.3 million shares outstanding as of this writing. A check will suffice, thanks.

We've seen one-time payments before. Microsoft offered shareholders one in 2004, $3 per share. Were Google to offer investors something similar today, it would create needed goodwill yet allow management the option to retain future earnings.

Here in the U.S., investors view dividends as a serious commitment, to be issued only when retained earnings are failing to produce above-average growth. Does that sound like Google to you? Yeah, me neither.