Jeff Bennett blogs at New Jersey Education Aid.
In the last two years there’s been a campaign by progressive groups in New Jersey such as Our Revolution, the Working Families Party, and the New Jersey Policy Perspective to eliminate the use of a “county line” in New Jersey’s primary elections.
What this campaign aims to abolish is the practice by almost every county’s Democratic and Republican Party organization to group its preferred local, county, and legislative candidates directly under the name of the candidates for the most prominent state & national offices, eg, Cory Booker, Robert Menendez, and Phil Murphy, plus their high-profile Republican counterparts.
The county line is impossible to defend as an ideal of an equal playing field within elections. The coronation of Robert Menendez Jr as the Eighth District’s new congressman, before anyone else even had a chance to declare their candidacy, exceeds my cynicism about American politics.
But while I agree that the existing county line system is bad, what is much worse is the paradigm of a partisan primary, which features a tiny, unrepresentative electorate and makes it very difficult for a true centrist or mixed-ideology candidate to be nominated and elected.
If you are a centrist who is discontented with both parties, the solution thus isn’t to abolish the county line for partisan primaries, but to abolish partisan primaries altogether in favor of a top-two or top-four “jungle primary” system.
If you want more diversity among New Jersey’s elected officials, we should not fragment candidate selection into hundreds of independent elections, but do what all other democracies do and centralize candidate selection.
I would not say that the current practice of the county line diversifies New Jersey’s politicians, but neither does it hinder it, and it’s the kernel of a mechanism that could produce demographically representative elected officials.
The Shallowness of American and New Jersey Democracy
I agree with critics that New Jersey has a weak, deeply flawed democracy.
Advocates say the county line is undemocratic because it gives party insiders more power since they are the ones who choose the candidate. They demonstrate that it’s unheard of for an incumbent to be defeated. Because the county line is so effective at steering votes, Julia Sass Rubin compares the “county line” to elections in the old USSR, where there was one name on the ballot and voting was a meaningless charade.
Anti-county line advocates blame the county line for the demographic non-representativeness of the New Jersey legislature:
The percentage of legislators who are male. NJ is in the bottom half of states in the US for gender equity in its state legislature [sic: this is not accurate]. Other groups are also sorely underrepresented in NJ’s legislature: Hispanics make up 7.5% of the legislature (20% of the NJ population overall); and Asians make up 2.5% of the legislature (10% of the NJ population overall). The Line is a barrier to entry for many underrepresented groups.
Opponents of the county line, however, don’t hide that they see the NJ Democratic party as insufficiently progressive and that their opposition to the county line is also motivated by wanting to pull the Democratic Party leftward. They point out the successes of Ayanna Pressley and Alexandria Ocasio-Cortez in defeating entrenched incumbents and want to make events like that more feasible in New Jersey.
They say party leaders are corrupt, anti-progressive, anti-woman, and anti-people of color. “It is especially heinous how the county line allows party bosses to shut down progressive women and people of color just because they put community needs ahead of the boss’s bottom line.” Imani Oakley, the Working Families’ challenger in District 10, says the fight to eliminate the party line is get people elected who “view progressivism as a tool of liberation and not just a convenient brand.”
I have many responses to this campaign.
First, what NJ does is unusual compared to other states, but not unusual compared to other democracies, which actually have much more diverse legislatures than NJ and the rest of the USA do.
I grant a license to exaggerate and I know that Julia Sass Rubin doesn’t actually think New Jersey is as undemocratic as the USSR, but parties in most other democracies don’t purport to let regular voters choose party nominees. In most other democracies, candidates are chosen by the party’s leaders. When regular people are allowed to participate, the “election” is more like a caucus and is something paid for by the party, not the government. In Australia they call the tiny number of people who chose nominees the “selectorate.”
It would be more literal to compare New Jersey’s system to Canada’s, which purports to let local voters decide party nominees, but there is so much rigging and ideological vetting that it is difficult for an outsider to even get onto a ballot, let alone be nominated.
Second, opponents of the county line juxtapose too closely the NJ legislature’s and House delegation’s demographic non-representativeness and the institution of the county line.
The anti-county line people are correct that New Jersey’s elected officials are non-representative demographically, but the New Jersey legislature is no more non-representative than in other states that don’t have county lines, so I am skeptical about the causality.
According to the Center for American Women in Politics, New Jersey’s legislature is 34.2% female, which is above the national average of 31.1%. The New Jersey legislature has a higher percentage of women than Massachusetts (30.5%), Pennsylvania (29.2%), and Delaware (30.6%). We are slightly ahead of Connecticut (33.7%) and only behind New York State (34.4%) by a tenth of a percent. The claim that NJ is below average in female representation is out of date.
As for race, according to a 2020 study by Politico, whites are overrepresented in the New Jersey legislature by 16.6 points, compared to by 12.4 points in New York, 16.1 points in Connecticut, 14.4 points in Massachusetts, and 13.6 points in Pennsylvania. NJ’s disproportionality isn’t nearly as bad as Delaware’s 23.7 point deficit in minority representation.
Nationally, New Jersey is slightly worse than the average, but comparing New Jersey to other states is not an apples-to-apples comparison. Fifteen states have term limits which make their legislatures younger and more diverse. Overwhelmingly white states like Vermont, Wyoming, Maine, have small gaps because their non-white populations are so small.
The NJ legislature’s racial disproportionality is unacceptable (an effective solution will appear below), but the disproportionality is inseparable from it being 20 years older than the average New Jerseyan.
The NJ legislature’s average age fluctuates between 56 and 59, depending on how recently the previous election was. New Jersey’s average age, by contrast, is only 39.
New Jersey is now nearly half non-white, but that doesn’t apply to the generation born before 1960 that rules the state. The county line theoretically makes it harder for younger candidates to win, but the gerontocratic nature of the New Jersey legislature is the same as other states that lack term limits.
According to Pew in 2015, the average NJ legislator was then 56, compared to 54 for Pennsylvania, 56 for New York, 58 for Connecticut, 59 for Delaware, and 54 for Massachusetts.
So despite our county line, we aren’t doing a worse job than other non-term limit states.
How New Jersey and all other states are glaringly non-representative in gender, age, and career is not against other states, but against the democratic nations that don’t even pretend to let local voters choose party candidates and just make candidate slates centrally.
I’ll look at age because it’s closer to being an apples-to-apples comparison than international racial demographic comparisons would be.
The average age of the Bundestag is 47, as is the lower house of the Spanish parliament. The average age of Mexico’s Chamber of Deputies is 48. Brazil is 52. The average member of the Italian parliament is a mere child of 44.
The youth of other countries’ legislatures is even more of a glaring contrast to the US and NJ because European countries are older than the US. Even, Japan, which is the world’s oldest country with an average age of 48, has a younger legislature than we do, with an average age of 55.
The youth of these other countries’ legislatures is not because they use proportional representation but because party insiders pick nominees and they want candidates to come from many walks of life.
France does not use proportional representation, but the average age of the National Assembly is 46.5. The United Kingdom doesn’t use proportional representation either, but the average age of the House of Commons is 49. Australia doesn’t use proportional representation for its House of Commons, but it has an average age of 51. Canada’s is 52.
These legislatures also have far fewer lawyers than the NJ legislature and US Congress. The US Congress is 43% lawyers; New Jersey’s legislature is 23% lawyers. In Europe, most countries’ legislatures are about 15% lawyers; in New Zealand they think it’s a problem that 17% of parliament is lawyers.
Central candidate selection as opposed to independent primaries also allows for (though does not guarantee) much greater representation of women because the insiders can intentionally create gender-diverse slates and quotas can be set. The dozens of countries that have higher representation of women than the United States do is because of centralized selection. Quotas for female representation are inconceivable if each nominee is chosen independently in an American-style, uncoordinated primary election system.
So if we want a younger legislature with professional, gender, and racial diversity, it’s better to let the parties choose the nominees, not each district’s voters. In other words, the practice of letting party leaders choose the nominees is what could create diversity.
As for NJ’s House delegation, it badly underpresents women (2/12) and underrepresents Latinos (1/12 vs 2/12 at proportionality), but it has the same issue of being very old, with an average age of 63. After Rep. Bill Pascrell retires he is likely to be replaced by a Latino and the NJ House delegation will become racially proportional at that point. If you include NJ’s two senators, New Jersey’s Congressional delegation is racially proportional.
The Bosses or the Bases?
Where I think proponents of county line abolition are correct is that the county line system makes it harder for anti-establishment progressives. Whether this is good or bad is subjective, but making it easier for anti-establishment progressives and their conservative Tea Party/MAGA counterparts in the GOP invites wider polarization.
But why have partisan primaries at all?
My problem with the anti-county line campaign is that it implicitly accepts the use of partisan primaries, despite them having multiple undemocratic features of their own.
First, partisan primary electorates are tiny and cede electoral power to the “third of a third” that turns out.
In NJ’s last gubernatorial election with two contested primaries, 2017, only 750,000 people voted – 500,000 for the Democrats and 250,000 for the Republicans, a minuscule 12% of registered voters for the two parties combined.
Turnout is even lower in off-off-year elections (ie, 2023, 2019, 2015 etc) and when one party does not have a contested gubernatorial primary. For instance, in 2019 the primary turnout was only 8%. In Passaic County it was 5%. Since nearly all legislative and House districts are non-competitive in general elections, a very tiny section of each district’s electorate determines winners.
Turnout is also extremely low in NJ primaries for the federal midterms. Turnout in 2018 was 9.8%. Turnout in 2014 was 8%.
2020 was a presidential election year and primary turnout soared to 26%, but 2020 was an exceptional year.
The American states that use jungle primaries for state-level elections have MUCH higher turnout.
- 2.5 million people voted in Washington State’s 2020 gubernatorial jungle primary which equals 51% of Washington State’s registered voters. Turnout was 34% in Washington State’s 2018 Senate jungle primary.
- Louisiana’s turnout in its 2019 jungle primary was also immense: 1.34 million compared to 3 million registered voters, or 45%.
- Finally, California uses a jungle primary and likewise has a turnout that dwarfs New Jersey’s. In 2018, 6.8 million Californians voted in their gubernatorial jungle primary and almost as many for its Senate jungle primary, which equals 37-38% of registered voters.
Political scientists have tracked California, Louisiana, and Washington and also found that their House of Representatives members are more moderate.
Second, primary voters are ideologically unrepresentative. They disproportionately consist of high-information voters who are farther right and farther left than general election Republican and Democratic voters, not to mention farther right and left disaffected centrists who don’t vote. The evidence is inconclusive about how much farther to the right and left the Republican and Democratic primary electorates are than general election electorates; political scientist Lee Drutman says they are only slightly more extreme, but no one thinks they are representative either.
However, there is evidence that elected officials have an “outsize fear” of Base primary challenges. Hence partisan primaries create a polarizing dynamic, pulling Republicans to the right and Democrats to the left.
The progressive opponents of the NJ county line lament how it is difficult for a young progressive to get a Democratic nomination, but it would be much more difficult for a median-voter centrist or mixed-ideology candidate to be nominated because centrist and mixed-ideology voters are split between the Republican and Democratic parties, whereas voters of the right and left are concentrated in the Republican and Democratic parties, respectively.
Since New Jersey uses semi-closed primaries, even if a Republican or Democrat voter wanted to pragmatically support a moderate candidate of the other party, they would not be allowed to vote anyway unless they went through the trouble of changing their party registration weeks in advance.
Politicians know that cross-party appeals are useless in a closed primary, hence they don’t attempt that as a strategy (even if they sincerely agree with the other party on one issue) and stay in lockstep with the rest of their party.
Alaska Leads the Nation
Again, not all states have partisan primaries.
California, Washington, Louisiana, and Nebraska use jungle primaries, though Nebraska’s runoffs are technically non-partisan and are only for the legislature, not the governorship or federal positions. Maine and Georgia use runoff or Ranked Choice Voting for certain elections, but they retain partisan primaries.
It is Alaska, however, that really is a beacon because it has combined primary reform with Ranked Choice Voting.
Alaska voters simultaneously abolished partisan primaries and implemented Ranked Choice Voting in 2020’s “Ballot 2.” Under Alaska’s new electoral law there is a jungle primary that advances the top four vote-getters, who then have a Ranked Choice Voting (RCV) general election.
Alaska’s reform disempowers the party bases since they cannot prevent a centrist or heterodox candidate from getting into a general election. With four slots in the general election and then RCV weakening the spoiler effect, it gives viability to true centrists and mixed-ideology candidates who lack it everywhere else.
Senator Lisa Murkowski is conservative, but she is pro-choice, she voted to convict Donald Trump, against the confirmation of Amy Coney Barrett, and now for the nomination of Ketanji Brown Jackson. In 2021, she was censured by the Alaska Republican Party, who vowed a primary challenge, which indeed has occurred from Kelly Tshibaka. Kelly Tshibaka is a Trumpist candidate, but is even more strongly anti-abortion and anti-gay rights than Trump himself.
If Alaska were still using a partisan primary and Murskowski were only accountable to Alaska’s Republican voters, Murkowski’s primary defeat would be all-but assured and she would have to attempt a write-in campaign (like she miraculously pulled off in 2010). However, due to Alaska’s top-four RCV electoral system, Murskowski can make it into a general election where she has a credible chance.
(Ranked Choice Voting is much better than the plurality-winner status quo, but a newer type of single-member district voting known as STAR voting is superior in electing consensus candidates to RCV because it simultaneously counts first, second, and third place preferences.)
But Why Keep Single-Member Districts?
Some elections in New Jersey, like for governor, Senator, and House of Representatives, must be single-member, but it would be legally possible for New Jersey to join the rest of the democratic world and adopt a form of proportional representation for its own legislature.
Single-member districting allows gerrymandering. Even when states district by bipartisan commissions, single-member districting is prone to disproportionate results where party seat representation can differ significantly from its share of the vote, as happens in NJ and most other states. Single-member districting even allows artificial majorities where the losing party ends up with more seats, as happened in NJ in 2013 and 2009 and the US House in 2012.
Proportional representation would ensure that everyone’s vote counts equally since districts are so large and heterogenous that differences in turnout would be small. Most importantly, it would allow alternative parties to gain representation, from Bernie Sanders-type Democrats to MAGAs to centrists.
Under proportional representation, urban center-right candidates would be elected and rural center-left candidates would be elected, giving each side of the spectrum an incentive to target all voters, not just a handful of exurban swing districts. I believe a voting system where all votes in all places count equally would lead to better government, especially in how it would incentivize Republicans to care about urban policy.
Single-member winners are especially absurd for the New Jersey Assembly, since it’s very common for one party to win each election narrowly and then claim both seats in districts that are essentially at parity.
If you had multiple viable options for your general election vote and each party’s slate was more demographically representative of the USA and New Jersey, would you care if the nominees are chosen by a party’s county leadership and not local voters? I wouldn’t.
Neither do people in Canada and Britain. There are mainstream movements for proportional representation in both countries – in 2015 Justin Trudeau sensed the popularity of election reform and promised voting reform – but fewer Canadians and Britons care about their functional inability to select party nominees.
The Catch-22 of Structural Reform
The problem for reformers in New Jersey is that the undemocratic features of the NJ government, such as partisan primaries or the county line, are protected by those features themselves. Our legislators and governors are elected through our plurality-winner primary, plurality-winner general election system, and hence they have no desire to make their own seats competitive or share power with other parties. Parties only favor structural reform when they are weak, and when they are weak they cannot enact those reforms.
State Senator Andrew Zwicker and Linda Greenstein actually do support Ranked Choice Voting, but their bill has attracted few cosponsors and Phil Murphy does not support it, though he does not candidly say so.
The recent vote reform states–California, Washington, Maine, and Alaska— all passed their runoff and Ranked Choice Voting reforms through initiatives. Nebraska adopted runoffs in an initiative back in 1934. New Jersey lacks initiatives and thus voters have no way to bypass the legislature and governor.
New Jersey has a Catch-22: an initiative process is necessary to change the structure of the government, but an initiative process cannot be created from outside the government itself.
Down with Partisan Primaries
The campaign against the county line is motivated by seemingly noble ideas about democratizing New Jersey’s elections, but the bigger democratic shortcoming of partisan primaries is the small, unrepresentative electorate itself and then, as an exacerbating secondary problem, NJ’s use of semi-closed primaries.
In a jungle primary > Ranked Choice Voting system I think including endorsements on the ballot would be helpful as a label to less-engaged voters, who might want to know who the mainstream county committee is backing, or who is a “Make America Great Again” Republican, or a “Working Families” Democrat.
And centralization of candidate selection seems to lead to greater diversity, not less. The problem with the county line is that it’s done at the county level, when perhaps it would be better at the state level.
Fragmentation of nomination leads to an old legislature with overrepresentation by whites, men, and lawyers. If we had more than one viable choice, either through a top-four general election or proportional representation, it would be irrelevant to me if the nominees were chosen by insiders.
Ranked Choice Voting would be a substantial reform, but a Ranked Choice Voting general election whose candidates are produced by partisan primaries and is Republican versus Democrat versus minor parties would not be that different from the status quo. If New Jersey retained its “sore-loser law,” a centrist or mixed-ideology candidate with broad appeal who lost the Republican or Democratic primary to a Base candidate would not be able to compete in a RCV general election anyway.
Thus, to really give different types of candidates a chance if we are going to keep single-member elections, Ranked Choice Voting (or STAR) and top-four primary reform should be combined, as Alaska has done.
The advocates for abolishing the county line constantly associate the county line with party leaders/bosses, but if you are a centrist or mixed-ideology person, is the influence of party bosses that much worse than party bases?
Although I’ve been critical of the anti-party line reformers in this op-ed, I respect them for the research they’ve done and for the awareness they’ve raised. I agree that New Jersey’s elected officials are unacceptably non-diverse, but I think the solution to that is to consolidate the nomination process. I agree that New Jersey, like the United States as a whole, is not a full-fledged democracy. As a centrist I believe the problem with New Jersey’s primary elections is that they are polarizing partisan elections.
