POLITICS
American Politics In The Networking Era
By Michael Barone, National Journal
© National Journal Group Inc.
Friday, Feb. 25, 2005
This article is excerpted from The Almanac of American Politics 2006, which will be published this summer.
On the surface, the 2004 election looked very much like the 2000 election. George W. Bush was again running against a liberal Democrat who had spent much of his career in the Senate and who had clinched his nomination by early victories in Iowa and New Hampshire. In November, 47 of the 50 states and the District of Columbia voted for the candidate of the same party as they had in 2000. Only three states switched, New Hampshire to the Democrats, Iowa and New Mexico to the Republicans. Bush won again, this time without a court battle. Republicans ended up with majorities in both houses of Congress. But in many ways, the 2004 campaign was very different from 2000. It produced a different kind of politics, a politics that reflects the character of the post-industrial, networking age we live in.
Changes in politics resemble changes in the larger society. For several decades now, we have seen the change from industrial America to post-industrial America, from an industrial nation characterized by centralization and large command-and-control organizations to a post-industrial, Information Age nation characterized by decentralization and network-connected organizations.
This is an America where Microsoft overtakes IBM, where FedEx overtakes the U.S. Postal Service, where Wal-Mart overtakes Sears. It is an America whose network-connected Special Forces overthrow the Taliban in Afghanistan and whose network-connected Army and Marines overthrow Saddam Hussein in Iraq. It is an America where the abolition of guaranteed welfare has produced higher incomes and greater independence for the target population, where network-connected police forces have cut crime by more than half in New York City and shown the way toward vast reductions in crime across the nation.
Our private sector and important parts of our public sector have moved from industrial command-and-control America to post-industrial, Information Age, network-connected America. In 2004, our politics followed.
The Politics of Networking
This was an election whose outcome cannot be dismissed as a fluke. It was an election in which most voters and most partisan activists on both sides believed that big things were at stake. It was an election in which the outcome by no means seemed certain: John Kerry led in the polls during a good portion of the time from March 2, when he clinched the Democratic nomination, to Election Day, November 2. It was an election that both sides believed would be determined by turnout. They had good reasons to think so.
The 1990s saw a decline in ticket splitting and a convergence of the two parties' percentages in presidential and congressional voting. Bill Clinton was re-elected with 49 percent of the vote in 1996, while the popular vote for the House that year was 49 percent Republican and 48.5 percent Democratic. In 1998, the popular vote for the House was 49 percent Republican and 48 percent Democratic. In 2000, both Al Gore and George W. Bush won 48 percent of the vote, while the popular vote for the House was again 49 percent Republican to 48 percent Democratic. The House vote in 2002 was a little different, 51 percent Republican and 46 percent Democratic.
But polling in late 2003 and for most of 2004 indicated a very close presidential race. Bush strategist Karl Rove keeps a card in his pocket showing that the percentage of voters who were behaviorally "independent" declined from 15 percent in 1988 to 7 percent in 2002. The strategy that Rove designed and that Bush-Cheney '04 campaign manager Ken Mehlman executed was geared not to persuading the undecided and weakly committed voters, but to turning out the maximum number of Republicans. The Kerry campaign and other Democrats likewise saw their main task as turning out the party faithful.
Both parties succeeded. Total turnout increased by 16 percent -- a historic increase. In 2000, 105 million Americans voted; in 2004, 122 million did. Turnout as a percentage of eligible voters increased from 51 percent in 2000 to 61 percent in 2004 -- again a historic increase. This was particularly extraordinary because turnout usually doesn't rise in rematches. In 1956, when Dwight Eisenhower ran against Adlai Stevenson for the second time, turnout increased only slightly; it decreased as a percentage of eligible voters. In 1996, when Bill Clinton faced a decorated World War II veteran for a second time, turnout declined in total numbers and, even more, as a percentage of eligibles. But in 2004, in the second contest between George W. Bush and a Democrat who had served in Vietnam and spent much of his career in the Senate, turnout zoomed upward.
But if both parties succeeded in raising turnout, one party was more successful than the other. Kerry won 16 percent more votes than Gore did. Bush won 23 percent more votes in 2004 than he did in 2000. The number of voters for the Democratic nominee increased from 51 million to 59 million. The number of Bush voters increased from 50 million to 62 million.
The parties went about raising their turnout in different ways. The Democrats depended on labor unions, as they had in the past, and on the turnout efforts of billionaire-funded "527" organizations. (These are named after a section of the Internal Revenue Code, and a number of these groups were funded by rich men like George Soros, who spent $27 million trying to defeat George W. Bush. Thank goodness the McCain-Feingold campaign finance law got the Big Money out of politics.) These groups relied on paid workers supervised by command-and-control organizations. They concentrated on black neighborhoods in central cities and on university towns -- areas where new voters would likely vote 90 percent Democratic. This was traditional, industrial-era politics, well executed. The Democratic groups met their turnout goals and more; if they had performed this well in 2000, Gore would have won by a large margin.
The Bush campaign was different. Its architect was Rove, who remained in the White House and advised President Bush on policy as well as politics; its structural engineer was Mehlman, who created an organization unlike any seen before, a networking organization that far surpassed what the Democrats were doing.
In mid-2003, when former Vermont Gov. Howard Dean surged ahead of other Democrats in fundraising and in the polls, much attention was given to campaign manager Joe Trippi's use of the Internet. He used it to bring volunteers and money into the campaign, and to allow Dean supporters to add their own words, literally, in the campaign blog. Many political supporters were impressed, and rightly so, that the Dean campaign amassed a list of 600,000 e-mail addresses. But few reporters at the time took note of the number of e-mail addresses the Bush campaign had collected: 6 million.
Over two years, the Bush campaign built an organization of 1.4 million active volunteers. This was unprecedented. By way of comparison, the Democratic National Committee has said it enlisted 233,000 volunteers during the 2004 campaign. The Bush volunteers worked not just in heavily Republican neighborhoods -- only 15 percent of Republican voters, Mehlman calculated, live in precincts that vote 65 percent or more Republican. Instead, they went everywhere, especially to rural counties, many of them slow-growing places where most politicians figure there are no more votes to be won, and to the fast-growing exurban areas at the edges of metropolitan areas, where most of the young families moving in tend to be Republican. Just as Sam Walton figured he could make huge profits selling things to people in low-income rural areas and in low-fashion exurbs, so Mehlman calculated that he could wring votes out of areas that most political strategists and political reporters ignored.
To make sure that those volunteers were achieving their goals, Mehlman established metrics -- numerical goals, measured by third parties. Every week, the leaders of the local, state, and national organizations got reports on whether those metrics had been achieved. Productive volunteers were given positive reinforcement, sometimes a call from Mehlman himself. Unproductive volunteers were replaced or persuaded to do more. Mehlman's management was very much like former Mayor Rudolph Giuliani's management of the New York City Police Department: Precinct commanders were given goals -- low crime numbers -- which were independently validated. Those who produced were promoted; those who failed lost their jobs. As a result, crime in New York was cut by more than 50 percent -- more than even Giuliani thought was possible.
This is not command-and-control management, but management by networking, by holding people accountable and letting them learn from each other how to do better. And in post-industrial America, it got better results than command-and-control management. In crucial states with the largest volunteer organizations, the numbers speak as loud as Giuliani's -- turnout rose 28 percent from 2000 in fast-growing Florida and 20 percent in slow-growing Ohio.
The Bush campaign used connections -- networks -- to recruit volunteers and identify voters. The campaign built on existing connections -- religious, occupational, voluntary -- to establish contacts. If a Bush volunteer was a Hispanic accountant active in the Boy Scouts, the campaign would reach out through him to other Hispanics, accountants and their clients, and Boy Scout volunteers. Of course, the campaign put much effort into contacting people in religious groups -- particularly evangelical Christians, but also Catholics and Orthodox Jews. And the Bush campaign reached out to people with shared affinities who tend to be Republicans. The campaign consulting firms National Media and TargetPoint identified Republican-leaning groups -- Coors beer and bourbon drinkers, college football TV viewers, Fox News viewers, people with caller ID -- and devised ways to connect with them.
As Thomas Edsall and James Grimaldi wrote in The Washington Post after the election, "Surveys of people on these consumer data lists were then used to determine 'anger points' (late-term abortion, trial lawyer fees, estate taxes) that coincided with the Bush agenda for as many as 32 categories of voters, each identifiable by income, magazine subscriptions, favorite television shows, and other 'flags.' Merging this data, in turn, enabled those running direct-mail, precinct-walking, and phone-bank programs to target each voter with a tailored message."
Presidential campaigns from 1968 up through 2000 spent most of their time, money, and psychic energy on devising television ads to appeal to undecided and weakly committed voters. Bush-Cheney '04 spent unprecedented amounts of time, money, and psychic energy on networking -- making connections with voters -- through advertising, to be sure, but also through personal contact. The Democrats' turnout drive depended on paid workers persuading strangers to get out and vote. The Republicans' turnout drive depended on volunteers persuading people with whom they had something in common to get out and vote. In industrial America, the Democrats' way may have been more effective. In Information Age America, the Bush campaign's strategy was more effective.
In his book Bowling Alone, Harvard professor of public policy Robert Putnam argued that America is suffering from a decline in social-connectedness -- in people voluntarily working and playing together, being active in those voluntary associations that Alexis de Tocqueville identified as one of the defining characteristics of democracy in America in the 1830s. The Bush campaign, by assembling a core of 1.4 million volunteers, increased social-connectedness in America in an important way.
Anyone who has volunteered and worked actively for a political campaign knows that it is a way to make new friends, to establish ties with people with whom you will work together again, on political campaigns but also on community projects and in voluntary associations of all kinds. Volunteer campaign work has reverberations over the years. Rove and Mehlman believed that it was possible to build such a large volunteer organization, but only for an incumbent president whom people had come to know well and admire, or even love. The Republicans will not have an incumbent to campaign for in 2008. But the 2004 Bush campaign created a quantum of social-connectedness that the Republican nominee in 2008 can build on, a long-lasting asset for the Republican Party.
Reshaping the Electorate
In the process, the Bush campaign reshaped the electorate. People who have voted once are more likely to vote than are people who have never voted. The Bush campaign added more people to the electorate in 2004 than the Democrats did, and that achievement is likely to reverberate in elections to come. It could even lead to the kind of natural majority for the party that the Democrats built in the 1930s and the Republicans built in the 1890s, majorities that pretty much prevailed for more than 30 years.
Recall that total turnout increased 16 percent in 2004. This was extraordinary: Increases in turnout in presidential years have averaged only 7 percent in the last 110 years, if one puts to one side the larger increases in the years from 1916 to 1928, when women were entering the electorate for the first time. Massive increases in turnout have the potential to reshape the electorate and create a new majority for a party that adds many more new voters than its rival. Over the 110 years preceding 2004, again disregarding 1916 to '28, turnout increased by more than 14 percent in only four elections. Two of those four elections resulted in a new national majority for the winning party.
The two that did not produce natural majorities were in 1992 and 1952. Turnout in 1992 was up 14 percent from 1988, and Clinton won this three-way race with 43 percent of the vote. That gave Clinton a chance to build a majority party. But this politically gifted president failed to do so. After his first two years, and in response to his tax increase and to the health care finance plan promoted by Hillary Rodham Clinton, Republicans won majorities in the Senate and, for the first time in 40 years, in the House of Representatives. Clinton failed to get 50 percent of the vote in 1996, and Gore, his vice president, failed to do so in 2000. Democrats did not win 50 percent of the vote in the House elections of 1994, 1996, 1998, and 2000. And these were all years of apparent peace and apparent prosperity, the most favorable atmosphere in which an incumbent party can run. In a counterfactual world without Ross Perot or Ralph Nader, Clinton and Gore might have won majorities in 1996 and 2000 -- but very small majorities, and we don't live in a counterfactual world. Now, the Democrats will not be able to run as the incumbent presidential party in a time of peace and prosperity until 2012 at the earliest.
The year when turnout increased the most was 1952, when it rose 26 percent over 1948. That was the year in which the G.I. generation entered the electorate in great numbers. Over the preceding 12 years, they had been fighting a war, moving around the country to work in defense industries, going to college on the G.I. Bill, trying to scrape up money for a down payment on a house. Then, in 1952, faced with the choice of two attractive candidates after 20 years of Democratic presidents, they started voting in a rush. The 1952 turnout produced a victory for Dwight Eisenhower, but he made no effort and apparently had no inclination to build a natural majority for his party.
That was not true of Franklin Roosevelt. He won re-election in 1936 over Alfred Landon by 61 percent to 37 percent, in an election in which turnout rose 15 percent from four years before. Roosevelt won 22 percent more votes than he had in 1932 -- almost exactly the same percentage by which George W. Bush increased his vote take from 2000 to 2004. Roosevelt reshaped the electorate, adding millions of voters in the big cities and cementing them to his Democratic coalition of Western progressives (mostly Republican in the 1920s) and Southern whites (solid Democrats until the 1950s) that became the dominant force in American politics for a long generation.
Roosevelt reshaped the electorate, in part, by adopting new political tactics appropriate to an industrial America caught in economic depression. Democratic machines had not previously dominated the politics of all big cities -- Chicago had two-party competition, and Philadelphia was solidly Republican in the 1920s. But Roosevelt's welfare programs and his cultivation of machine politicians, combined with the popularity of his New Deal in the cities, meant that the machines, sometimes with helpful subsidies from government, could roll up huge majorities simply by turning out anyone they could find (and perhaps some they couldn't) in now monopartisan constituencies. In addition, the labor law that Roosevelt signed in 1935 permitted the rapid growth of industrial unions starting in 1937 and continuing through World War II, and those unions became powerful turnout machines for the Democrats. Roosevelt reshaped the electorate and made it predominantly Democratic for a long generation.
The other instance was the election of 1896, when turnout increased 15 percent over 1892, and Republican William McKinley was elected over populist Democrat William Jennings Bryan by 51 percent to 47 percent. McKinley received a startling 37 percent more votes than the incumbent Republican president, Benjamin Harrison, had four years before. Mostly, this election is remembered for Bryan's brilliant oratory and his fervent advocacy of "free silver" -- in effect, inflation -- as the savior of the farmer. But McKinley's platform of hard money and protective tariffs had more appeal to workers and immigrants in the growing big cities. McKinley's campaign, managed by Sen. Mark Hanna of Ohio and future Vice President Charles Dawes, pioneered new tactics to appeal directly to the big- and small-city masses and won votes the Republicans had never won before. McKinley did not win big -- his advantage was almost identical to that of Bush over Kerry in 2004 -- but he did succeed in reshaping the electorate and setting forth policies that would continue to increase Republican support in the years following. For a long generation, until the Depression of the 1930s, the Republicans were the natural majority party in the country.
Interestingly, neither McKinley nor Roosevelt was considered a sure winner during the fiercely waged campaigns of 1896 and 1936 -- another similarity to 2004. Bryan was the candidate of the party in the White House, a party that had won more popular votes for president in four of the five preceding elections, and his oratory and appeal to farmers seemed likely to add more. Few foresaw how the push by Hanna and Dawes to recruit voters in the cities and factory towns would reshape the electorate. As for Roosevelt, he faced a mostly hostile press, and many of his critics doubted that he could hold together what seemed to them an unwieldy coalition of opposites. Republicans, after all, had won big majorities in three of the four preceding elections. Random-sampling polling had just been developed: George Gallup published his first scientific poll in October 1935, and his later surveys showed Roosevelt leading. But many observers had more faith in the familiar Literary Digest mail-in poll, which showed Roosevelt trailing Republican Landon.
The 1896 and 1936 elections reshaped the American electorate, in ways that produced natural majorities for the winning party. The winning parties succeeded, in part, by inventing new political techniques appropriate to the times. The 2004 election has also reshaped the American electorate, in part through the invention of new political techniques. It is too early to say that it produced a natural majority for the winning party. But it has laid the groundwork.
The votes in 1896 and 1936 were harbingers of long-lasting majority coalitions for their parties. Political strategist Rove, who has often identified McKinley and the coalition he built as a model, hopes that the vote in 2004 will be, as well.
Like Clinton's election with 43 percent of the vote in 1992, Bush's election with 48 percent of the vote in 2000 gave the winner an opportunity to build a majority for his party. So far, Bush has succeeded where Clinton failed. In the 2002 elections, Republicans ousted the Democratic majority in the Senate and increased their majority in the House. No incumbent president's party had increased its number of seats in both houses in an off-year election since Roosevelt's Democrats did it in 1934. In 2004, Bush was re-elected by 51 percent to 48 percent, and Republicans increased their margins in the Senate and the House; the popular vote for the House was 50.1 percent to 48 percent. Note that Republicans won these admittedly small majorities not in times of apparent peace and apparent prosperity, but in times when the nation was under attack and facing war, and when the economy was in recession or was in what was widely, though inaccurately, described as a jobless recovery. In other words, the Republicans were facing the voters in 2002 and 2004 in a more unfavorable posture than the Democrats had had in 1996, 1998, and 2000. The Democrats, in the more favorable posture, failed to win majorities. The Republicans, in a less favorable posture, succeeded in winning majorities.
In this space four years ago, we described America as "the 49 percent nation," evenly split between the two parties, with voters divided especially along lines of religious belief and observance. That division remains, but the numbers have changed. America is now, perhaps momentarily, or perhaps at the beginning of a long period, a 51 percent nation, a majority -- a narrow majority -- Republican nation. The evidence is there in Bush's re-election victory, in the Republicans' popular-vote majorities in the 2002 and 2004 elections to the House, in the 2004 National Election Pool exit poll that showed party identification at 37 percent Democratic and 37 percent Republican, compared with the 39 percent-to-35 percent Democratic advantage registered in the Voter News Service exit polls in 1996 and 2000. Clinton had the chance to forge a majority for his party. He failed. Bush had the chance to forge a majority for his party. He succeeded.
The Reshaped Electorate
The new shape of this enlarged electorate shows up in comparisons of numbers from the NEP exit poll in 2004 against the VNS exit-poll numbers in previous elections. In 2004, the adjusted NEP figures showed party identification at 37 percent Republican and 37 percent Democratic. This is in contrast to the Democrats' 39 percent-to-35 percent edge in 1996 and 2000 and their much larger party-identification advantages in the 1970s and 1980s. This was the first election in which Republicans achieved parity in party identification since the invention of random-sample polling in the 1930s. In other words, this was the most-Republican electorate any American under age 80 has ever seen. It was also a conservative electorate; 34 percent identified themselves as conservative versus 21 percent who said they were liberal. That's a 5-percentage-point increase since 2000 in the number of self-identified conservatives.
Much of the campaigning -- the candidates' appearances, the advertising, the organizational work -- was concentrated in the battleground states. In October, they numbered 13 (Colorado, Florida, Iowa, Maine, Michigan, Minnesota, Nevada, New Hampshire, New Mexico, Ohio, Oregon, Pennsylvania, and Wisconsin), as the campaigns winnowed the list by striking off states that seemed hopelessly out of reach. Nine of these states have had lower-than-average population growth since 2000, and four -- Michigan, Ohio, Pennsylvania, and Wisconsin -- have had highly visible losses of manufacturing jobs. Overall turnout in these battleground states was up 20 percent, and both Bush and Kerry increased their party's total votes over 2000 by 22 percent.
In fast-growing Florida, Bush turnout was up 36 percent and Democratic turnout up 23 percent. Bush's 537-vote edge in 2000 was transformed into a 381,000-vote advantage in 2004. In Ohio, which was plagued by manufacturing-job losses and Republicans' state tax increases, Bush faced less-favorable terrain. Democratic turnout went up 25 percent, and Republican turnout went up 22 percent -- enough for Bush to hold a state that he had carried in 2000 by 166,000 votes; he won Ohio by 119,000 votes this time. In the battleground states generally, and particularly in Florida and Ohio, Kerry needed to do better to win, and didn't.
The reason that Kerry needed to do better to win is that the states that the Democrats conceded as safe for Bush had more electoral votes, 213, than the states the GOP conceded to Kerry, 179. And in both sets of states, the same pattern prevailed: The Bush vote rose more than the Democratic vote. Bush's popular vote increased 23 percent in the safe Bush states and 22 percent in the safe Kerry states. Kerry's vote increased 15 percent over Gore's in the safe Bush states and only 12 percent in the safe Kerry states. Bush carried the safe Bush states by 56 percent to 41 percent in 2000, and by 59 percent to 40 percent in 2004. Gore carried the safe Kerry states by 56 percent to 39 percent in 2000, and Kerry carried them by 56 percent to 43 percent. In sum: The 2004 results showed the red states getting redder and the blue states getting less blue.
All of this means that as long as something like the current contours of support prevail, Republicans start off a future presidential campaign with more electoral votes on their "safe" list than Democrats and have the potential to add more states to their target list. Of the safe Bush states, only one, Missouri, with 11 electoral votes, went to the president by a result as narrow as 53 percent to 46 percent. In contrast, three safe Kerry states, Delaware, New Jersey, and Washington, with a total of 29 electoral votes, were that close.
Religion is, as pointed out in previous editions of The Almanac of American Politics, the great divider, the demographic variable that correlates with voting behavior -- and what matters is not only religious denomination, but also degree of observance. Pundits made much of Rove's claim that turnout of white evangelical Protestants in 2000 was 4 million less than expected and of his determination to get them to the polls in 2004. He probably did, but they don't seemed to have increased their percentage of the electorate much; they were 23 percent of voters in the 2004 NEP exit poll; the 2000 VNS exit poll didn't have the same category. Bush won 78 percent of their votes this time, and so they were indispensable to his victory; but so were others. Bush's percentage among all Protestants was up 3 percentage points, to 59 percent; among Catholics, up 5 points, to 52 percent (against the first Catholic nominee since 1960); among Jews, up 6 points, to 25 percent; among those with no religion, up 1 point, to 31 percent. Those who said they attended religious services weekly voted 61 percent for Bush, up just 1 point; those who said they never attended voted 36 percent for Bush, up 4 points. Protestant weekly churchgoers voted 70 percent for Bush, and Catholic weekly congregants, 56 percent; Protestants who attended less often voted 56 percent for Bush, and Catholics who did so voted 49 percent for Bush. So the reshaping of the electorate has slightly reduced the polarization along religious lines.
If there is a religion gap within the American electorate, there is also a marriage gap -- a gap that is far wider than the oft-touted gender gap. Married people voted 57 percent to 42 percent for Bush; unmarried people voted 58 percent to 40 percent for Kerry. Those who said they were gay, lesbian, or bisexual voted 23 percent for Bush, just 2 percentage points less than in 2000, despite his support for a constitutional amendment banning same-sex marriage.
But if the Bush campaign has reshaped the electorate and made it more Republican, the changing racial or ethnic composition of the country will, over time, tend to make it more Democratic. That is the argument made by Ruy Teixera and John Judis in their thoughtful, if optimistically titled book, The Emerging Democratic Majority. The results of the 2004 election among Hispanics, Asians, and African-Americans do not provide much support for their thesis. To be sure, the leaders of Hispanic and Asian groups, like the leaders of black groups, are overwhelmingly Democratic. But the Bush campaign evidently was able to establish connections -- create new networks -- to reach Hispanic and Asian voters.
Hispanics are potentially the fastest-growing segment of the electorate -- 13 percent of the total population, 17 percent of the population under age 18, between 6 percent and 8 percent of the 2004 electorate. Bush and Rove, from their days in Texas, have been targeting Hispanics for conversion. Democrats have hoped that Hispanics would become, like African-Americans, a solidly Democratic voting bloc, so that registration and turnout drives among Hispanics would produce a reliable harvest of Democratic votes. Analysts dispute the actual Hispanic vote in 2004, but taking into account various exit polls and surveys, it seems prudent to estimate that Bush won about 40 percent of the Hispanic vote in 2004, about a 5-percentage-point increase from 2000.
National NEP data show Asians voting 44 percent for Bush, up 3 percentage points from 2000. In the two states with the largest numbers of Asians, NEP shows them voting 48 percent for Bush in Hawaii -- where Filipino- and Japanese-Americans have long shown a tendency to leave the Democratic Party to vote for incumbent Republican presidents -- and 34 percent for Bush in California, not far off his percentages in the San Francisco Bay Area and Los Angeles County, where most of California's Asians live.
The picture is quite different for African-American voters. Nationally, NEP showed them voting only 11 percent for Bush, up just 2 percentage points from 2000, although it does seem likely that the Democratic Party won't be able to count on the same 90 percent level of support from blacks that it has come to take for granted since the election of 1964. In any case, blacks are not a growing segment of the electorate, as Hispanics and Asians are and will continue to be for some time. And Hispanics and Asians are not lining up with African-Americans as a single, overwhelmingly Democratic bloc of people of color. Something rather different is going on, something more like the political progress of the immigrants of the period from 1840 to 1924. |
