High dividend stocks: Do they beat the market?

I was browsing through the Wall Street Journal this weekend and came across this story about "high dividend" stocks:
http://bit.ly/b0MrBt
Briefly summarizing, the author argues that investing in five high dividend paying stocks is a better strategy for an investor than investing in an index fund, and that the "loss of diversification" is made up for by the higher returns generated on the dividend paying stocks.

It may be surprise you, but I don't disagree with the core of this strategy, which is not a new one. In fact, in what is known as the "Dow dogs" strategy, you invest in the five highest dividend yield stocks in the Dow 30. A more detailed sales pitch for this strategy can be found here:
http://www.dogsofthedow.com/
Over time, its proponents argue that the strategy would have paid off richly for investors.

To add even more ammunition to dividend seekers, studies over the last three decades have also shown that stocks in the top decline in dividend yield generate about 2-3% higher returns, after adjusting for risk, than the rest of the market. This newsletter nicely summarizes the evidence:
www.tweedy.com/resources/library_docs/papers/highdiv_research.pdf

So, is this a winning strategy? In my book, "Investment Fables", I have a chapter on this strategy which you can download by clicking below:
http://pages.stern.nyu.edu/~adamodar/pdfiles/invfables/ch2new.pdf
Briefly summarizing, I argue that as with all investment stories, there are caveats and these are the most significant ones for "dividend" heavy strategies:
1. Dividends are not legally binding: Unlike coupons on bonds, where failure to pay leads to default, companies can cut dividends without legal consequence. The fact that "dividends are sticky" and companies don't usually cut dividends does not take away from this point.
2. Higher tax liability: At least in the United States, for much of the last century, dividends have been taxed at a higher tax rate than capital gains. (Since 2003, the tax rates on the two have been the same, but that law is set to expire on December 31, 2010, returning us to the old tax laws).
3. May be "liquidating" dividends: When companies are in decline, they may pay large liquidating dividends, where assets are sold to fund the dividends. While these dividends are cash flows, they are not sustainable and will run out sooner rather than later.
4. Sector concentration: If you pick the highest dividend yield stocks across a market at any point in time, you may find yourself holding stocks in one or two sectors. In early 2008, for instance, you may have ended up with five banks in your portfolio. If there is a shock to that sector, your portfolio will collapse.
5. The market "knows" something you do not: Remember that the dividend yield for a stock shoots up almost always because the price drops, not because the dividend is increased. In other words, it is a sudden drop in the stock price that makes the stock look attractive. It is true that markets make mistakes, but it is also true that sometimes price drops of this magnitude occur because there is a serious problem looming on the horizon.

Here is how I would modify the strategy to protect myself against these three issues:
1. Look for companies with positive earnings, low debt burdens and high cash balances. They will be under less pressure to cut dividends.
2. Use this strategy for "tax protected" portions of your portfolio. Even in the US, investments made in pension plans are allowed to accumulate income, tax free. Even if you cannot pick individual stocks in your pension fund, you may be able to redirect the money to a high dividend yield mutual fund.
3. Steer away from companies with dividend payout ratios that exceed 80% and have negative revenue growth. That may help keep liquidating companies out of your portfolio.
4. Try for some sector diversification. In other words, classify companies at least broadly into sectors and look for the highest dividend yield stock in each sector, rather than across the whole market.
5. Check every news source that you can find for news stories or even rumors about the company. 

In short, buying high dividend yield stocks makes sense for a long-term, tax-advantaged investment. One point that I disagree on is that this strategy requires you to give up on diversification. I don't see why you cannot construct a reasonably diversified portfolio (of 30-40 stocks spread across sectors) of high dividend yield stocks.