You can arguably hold either a full index of stocks and do the 'sell-off' at 3-5% per year, or you can pursue dividend-growth-investing (DGI) and live off only the dividends.
People on here seem to be providing pretty reasonable answers thus far and it all depends on 2 things:
1- Do you believe in the EMH? For a forum of strong self-aware savers and investors it seems EMH is the 'kool aide' here. That's fine. To each their own. EMH works fine if you can forget the 90s tech bubble and the 2008 housing crash and instead focus on y/y indexing averages. I'm a huge proponent of indexing myself-- most active managers suck at their job-- but I get a bit cringed when people try to argue that "all stocks are 100% efficient." Something like Tesla (TSLA) has a lot more assumptions priced into it than something like Coca Cola (KO), the risks are nowhere near equal, and beta is a poor proxy of (long-term) 'risk,' but now I'm digressing.
2- Do you plan a 'flat spending curve' or an 'upward spending curve?' Or, are you semi going to full retired or just flat/full retired?
If you believe in #1 it really doesn't matter what you do as long as you diversify enough (bonds, stocks, different industries) to shield yourself from outlier events. Modern Portfolio Theory (MPT) is your sauce here.
However, if you don't fully buy #1, and depending on your age, you probably want to stick to more stable names.
Also if you don't buy #1, then #2 becomes a real question.
If you want your income to ramp up, many investors like the DGI approach because they can live off only the dividends. Building a portfolio of 20-30 stable dividend payers (but also growers) provides a steady stream of income, usually around 2.5% if you add enough growth (stuff like Disney with lower yield but high growth) and avoid risky high-yielding instruments (stuff like Annally (NLY), shipping stocks) or high-yielders that will never/hardly grow the underlying (MLPs, and a lot of REITs fall in here).
The 'fun' part of DGI versus a 4% drawdown is that say with $1M, you start with $25k a year to spend. However, the average growth of these dividend stocks is going to be close to 7%. Next year you have $27k, in 10 years you have over $49k per year to spend. In 20 years you are spending over $97k. Since the 2.5% yield will probably stay around constant, in 20 years your actual wealth will be close to $3.9M. Meanwhile your 4% drawdown neighbor is now able to spend around half of what you do and his/her nestegg is $1M at best, probably smaller.
The clear thing to realize is that it's not the dividends that is the 'magic' necessarily. You can do the same with withdrawals. By doing 2.5% each year instead of 4-5%, your underlying portfolio will be getting bigger faster than you are taking the stocks out.
The reason dividends provide some 'magic' on the long-term is four-fold:
1) When stocks crash, the fixed withdrawal system decimates your base. For example a 5% withdrawal in spring of 2009 killed off well over 11% of your wealth by now. Dividends keep you focused on the payouts, which if you are with MCD, KO, WMT, TGT, etc, actually went UP in 2008-2010. Your base remained fully intact.
2) Dividend paying stocks often are stable industries and subject to less overall risks. The dividend payouts keep these companies from overexpanding or pursuing stupid acquisitions. These companies often have a core competency in which they focus.
3) Rules on your dividend portfolio allow you to quickly cut those who fail to raise their dividend, or worse cut it, which is a sign of bad things to come. Investors could have shed GM and Ford years before the crash, JCP would have been cut years before this fiasco, etc etc.
4) Dividends paying blue chips are often seen as 'safety' stocks. When the economy is having problems, investors flock to them. Often times the blue chips also provide goods that do well during a crisis. While they obviously suffer sales slowdowns in many cases, the slowdowns are a tiny piece of the broad economic hit. Look back at MCD or WMT stock during the recession- dividends up, profits barely hit, stock flat or even up.
