There's been quite a debate between free and paid content but it's in quite a sorry state. On one side (mostly old school reporters and newspaper execs) people think that publishers need to charge people for content, block everyone else out, and sue the pants of Google for profiting off indexing stories. On the other side (mostly techies and copyleftists) are free content idealists harping on the line that "information wants to be free."
But both sides are equally wrong, or at least misguided. Pretty much every company in the world has different products at different prices, why do journalists think they should have just one?
What you really need to do to monetize content is to take a page from the airlines and freemium web services and every other company with a product ever and learn how to price discriminate. Not only will you make much more money doing it, you will produce much better quality journalism as well.
Offer your basic and commodity content for free. Then create high value premium content and charge readers a monthly fee for it. 
Alan Mutter has some good ideas for types of premium content you can create. Here's the economics that underlies it.
The Demand Curve
We have a standard downward sloping demand curve  with each point on the curve indicates how much value a consumer is getting from the product. Now if a site gives away all their content, the only revenue they get is from advertising, and the consumers who get the most value from the content pick up a nice chunk of consumer surplus.
Point D represents the smallest amount of value a reader gets from your site. This is likely in the micropayment range, likely just a few cents, but the quantity consumed is also huge. It doesn't make sense to mess around with micropayments, because the site can monetize their traffic with advertising revenue. For simplicity's sake we'll say marginal advertising revenue is constant for any quantity.  While you could try charging with micropayments, it likely wouldn't work. The mental cost of deciding whether or not to pay and then paying is a hefty tax on the process. (I'd argue that trying to charge anything less than $10 at a time or so becomes counterproductive, but the smoother the payment process is the lower this amount can go)
How can you increase revenue? By capturing consumer surplus.
Price Discrimination with Advertising
We would like to set several different price points at A, B, and C to charge people different amounts based on where they are on the curve. Businessmen and politicians and news junkies get much more value from a news site than casual readers do, but how can we charge them a higher amount than everyone else? Airlines are the classic example. They do it with such a pricing schemes that make sure that you almost certainly did not pay the same amount for your ticket as your seatmate.  To do it for a news site, you'll want to create varied products such that the product offered at price point A appeals to all the readers left of point A. Each of the points A, B, and C need to be different slices of unique premium content that readers will pay different amounts for.
A few more points to consider
- Judging Value — You need to be able to accurately judge the value of all your content and spend a good amount of time researching and calculating what the proper price points are. I suspect for high quality business news, sites could get away with subscriptions costing upwards of a hundred dollars per year. Things like high end wine or restaurant reviews also don't make much sense as free content. If someone is spending a hundred dollars a bottle, they'll pay for a review backed by a renowned brand name.
- Subscriber value versus site traffic — For any given piece of content, you have a dilemma. You can put it in the premium content pile and restrict access to it, or you can put it up for free. Each piece of content that goes into the free pile increases your reach and traffic, but slightly erodes the value your premium subscribers get and makes them more likely to switch. This is a delicate balance. One way around it might be to delay the speed at which non-subscribers can access content. So that masterful investigative piece might be subscriber only for the first week, after which its open for anyone to read. The kind of high quality, high traffic pieces you want to grow traffic will be the exact same pieces of content you want as premium content to get people to pay.
- Archives are terrible premium content — A lot of newspapers have their full archives online behind a paywall. Sometimes a very expensive paywall where per article access can cost close to $10. What audience is this supposed to attract? People who place enough value on old news articles (lawyers, academics, other reporters, students writing papers etc.) to pay for them, likely all have access to LexisNexis subscriptions already and won't pay. And the average reader won't care enough to pay for an article they stumble across. It seems like they've made a terrible subscriber value/traffic trade-off here.
- What am I buying? — Transparency and convenience are important. With your premium content make it explicitly clear what types of content and privileges a subscriber will get. Any uncertainty here means fewer paying customers.
- Piracy will be an issue — Remember that when digital content has a price, piracy is going to be a problem. Look at the troubles the movie and music industries are having. You'll have to be willing to vigorously defend your copyright against sites which spring up and paraphrase all your premium content.
 DO NOT just repackage your current content as free content. It's almost certainly not good enough. Thinking hard about what sorts of content readers will pay for, you'll produce better content too.
 Different pieces of content have different value to different consumers. Together they form a downward sloping demand curve. A few consumers are willing to pay a lot for content, but this amount drops off quickly. On the right side of the graph, by decreasing the price towards zero, the quantity consumed can be expanded almost to infinity.
 In truth, marginal revenue from advertising is probably downward sloping as well, but the market for that is so much more inelastic than that for traffic that it doesn't matter. That line should also extend all the way to the left.
 Nearly every economics textbook has an example about the price of airline seats to explain price discrimination. For them point A are the business travelers that need comfy seats and need to buy them at the last minute. Because the value they get from travel is much higher than average they are willing to pay much more. Point B might be the families on vacation, they're willing to book well in advance but they might want their tickets refundable in case things change. But they're also price conscious. If the ticket costs as much as at point A, they'll just stay home instead. Point C would be the budget travelers and backpackers, who are willing to fly standby on whatever plane has a few empty seats. In the end they all get the same service, transportation from one point to another, but by varying the conditions and terms of each ticket, the airline is able to make much more money.
Looks like this post has gotten a bit of attention. If you found it interesting, you might also enjoy my earlier post 9 ways that newspapers can make money that aren’t advertising.
I've also given a brief talk on this topic at the Information Valet conference hosted by the University of Missouri's Reynolds Journalism Institute. Martin Langeveld did a great write up of it for the Nieman Lab, making some of my points better than I did myself. A video is here.