David Broder's years covering politics help make him one of the most knowledgable and serious commentators in journalism today. Along with Post colleague EJ Dionne, Broder may be the only regular columnist at a major newspaper whose books are on political science reading lists from Oxford, England to Oxford, Mississippi. Yet, sometimes this breadth of experience gets him in trouble.

Contra what some liberal commentators have recently argued, in today's Washington Post, Broder uses a recent Democracy Corps poll and memo to argue -- as its authors do -- that Democrats can't capitalize on Bush's unpopularity because they lack ideas. Not that they lack specific policy positions, but they lack the over-arching positive vision of where the country should go and how to take it there; a concept of how the world works - and how it should.

Broder astutely notes that Harry Reid and Nancy Pelosi are institutionally ill-suited for developing such a vision -- yet Howard Dean has outsourced this job to them. Broder then challenges Dean to have the "courage" to follow a different model, one found in the DNC's archives: that of the Democratic Advisory Committee, a group of elected officials and party leaders established in 1954 to lay out a long-term agenda for the Democratic Party.

To assess the impact that the Democratic Advisory Committee had on the Kennedy and Johnson years, I leave that to the expert on the matter, Phil Klinkner of Hamilton College. But, as a vehicle for change and idea-generation today, a party-based policy committee is the absolutely worst route to go.

First, the Democratic Advisory Committee is the wrong example to draw from for the simple reason that it did not operate in the modern era of party organization or campaigning. The party -- especially the political bosses and elected officials -- still had much more control over its agenda, and the discipline to have it carried out. More than that, the party platform actually mattered a whole lot more. Candidates looked to it for guidance; campaigns were actually run on it.

Now, we live in a era that is candidate-centered (that is, candidates raise money themselves, decide their agenda, and if they win the nomination dictate the platform of the party). Because of the need of candidates for volunteers and funds, organized interests have taken a central role in party affairs. And in the Democratic Party, in particular, securing a presidential nomination or drafting a platform has become mainly an exercise in interest-group management.

Broder has counseled Democrats to look in the wrong box of their archives (which, by the way, I last saw in the bowels of the National Archives around the corner from the Quayle Christmas cards). To see how (in)effective these bodies are, Democrats need to look not to 1955, but to 1985 and at a modern example of party-based policy-making: the Democratic Policy Commission (DPC).

The DPC was formed by newly-elected DNC Chairman Paul Kirk two decades ago. Like the Democratic Advisory Committee, the DPC was formed after the second bruising defeat of the party's presidential nominee and was populated mostly by elected officials.

The DPC, however, was hamstrung by the inherent fractiousness of the Democratic coalition, and in order to get a final document out, the commission had to avoid or water down certain issues. For example, the report -- entitled "New Choices in a Changing America" and released in the fall of 1986 -- ignored inflammatory social issues, especially abortion and gay rights. And an intense debate over military spending and the role of the US abroad delayed release of the final document for months. Even this careful bridging of the divides within the party wasn't enough. When the report was finally released, left-liberals of all stripes blasted it; and it quickly disappeared from view.

To Kirk's credit, the final report did embrace positions on economic growth and the Cold War that were laudable. But it had little impact. Let me ask you: have you ever heard of the DPC? I didn't think so.

Too many cooks with too many recipes, ultimately, can only whip up a large plate of warmed-over mush. If Dean was to undertake a major policy-making effort right now, he would -- like Kirk before him -- either incite a civil war (thus breaking a party chairman's first commandment: thou shalt make sure everyone gets along) or end up with a policy document that would have all the edge, bite, and imagination of a heap of mashed potatoes.

After the DPC faded from existence in 1986, another policy-focused group continued its work. They were able to come up with provocative proposals on anti-poverty policy, national service, and law and order (as well as an overarching worldview) that laid the foundation for the electoral and governing success of a young Democratic president. That group was the Democratic Leadership Council (rabid DLC-haters begin writing hateful comments now). And central to its success was that it was formed outside of the apparatus of the DNC.

That's the central lesson to learn from the DNC archives: in this day and age, effective policy-making happens outside of party control. What Dean needs to do is stay out of the policy-making business, and encourage donors and activists to fund new think tanks and fora for debate about what the Democratic Party stands for. Ideas matter. Parties need them -- they just can't be trusted to develop them.