Which way did Utah Co and Weber Co swing?

Trump’s 2020 gains are trivial. He continues to fare worse in these areas than any Republican nominee in a generation.

The New York Times notes in passing today that two Utah metro areas (Provo-Orem and Ogden-Clearfield) moved toward Trump in 2020–the 6th and 8th largest pro-Trump swings in the nation!

Trouble is, the NYT is correct only in a very limited sense. It is much more accurate to say that Democratic presidential candidates did better in these areas than any Democratic presidential candidate in at least a generation. (I’ll look at the county-level results for this analysis, since I don’t have metro-area results handy.)

The NYT’s approach: Raw ballots, 2020 vs 2016

The NYT is correct that these counties moved toward Trump only if we limit ourselves to (1) comparing 2020 to 2016 only, and (2) looking at the raw numerical increase (as opposed to percentage increase) in ballots cast for Trump versus Clinton-Biden. With these limitations, here’s what we see:

In Utah County, Trump won 90,629 more ballots in 2020 than 2016, while Biden won 47,511 more than Clinton. In Weber County, Trump won 25,714 more, but Biden won only 14,564 more. These numbers motivated the NYT’s conclusion that both areas swung toward Trump.

Even within the narrow frame of 2020 vs 2016–a problematic frame, but we’ll get to that later–I don’t love this conclusion. Let’s switch from raw numbers to percentages to see why.

Trump may have gained 91K additional votes in Utah County, but his increase from 102,183 votes (2016) to 192,812 (2020) was only an 89% increase, as opposed to Biden’s astounding 167% increase over Clinton’s share (from 28,522 to 76,033). Likewise in Weber County, Trump’s share rose by 64%, but Biden’s rose 76% over Clinton’s.

I know, I know. In an election, it’s raw ballots that win, not percentage growth. But the percentage growth tells you something important about the baselines each candidate built on. Relative to their starting points, Biden gained more ground than Trump.

Better: Ignore 2016 and look at the longer trend

McMullin made 2016 so weird in Utah that any comparison between 2016 and another presidential election will be fraught. McMullin won 30% in Utah County and 19% in Weber County. With him off the ballot in 2020, those votes had to go somewhere. If we really want to understand how Trump did here, we need to look at the longer series so we can see past 2016.

First, look at the vote share won by Republican presidential candidates since 2000 in these two counties. Utah County is blue, Weber is orange. Trump may have increased his share in both counties relative to 2016, but it’s clear he’s way off from a typical Republican candidate. Trump’s 2020 gains are trivial. He continues to fare worse in these areas than any Republican nominee in a generation.

For another perspective, look at each Democratic nominee’s vote share over the same period. Unlike the previous chart, 2016 doesn’t stand out this time, making it clear that McMullin mostly took votes from Trump, not Clinton, in 2016. In both counties, Biden won more votes than any Democratic nominee in a generation. His 27% in Utah County is huge–other than Obama’s much lower 19% in 2008, no other Democrat has broken 14% there this century.

The point

In a narrow, short-term sense, it is correct that these two parts of Utah shifted toward Trump relative to 2016. More accurately, though, Trump continues to underperform. Without McMullin on the ballot, Utah Republicans may have been more willing to hold their noses and vote for him. But with McMullin gone and turnout way up, Democrats posted their best numbers in these two counties in a generation.

Don’t forget that turnout also rose tremendously–by 41% (83K ballots) in Utah County and 31% (26K) in Weber County. We’re not looking at the same voters year-to-year. But that’s a topic for another day.

About Adam Brown: Adam Brown is an associate professor of political science at Brigham Young University and a research fellow with the Center for the Study of Elections and Democracy. You can learn more about him at his website.

Posted in Everything | Tagged , , , , , , , , , | Comments Off on Which way did Utah Co and Weber Co swing?

New data show how little time the Legislature spends per bill

The median bill enacted in 2018 received a total of only seven minutes of House and Senate floor debate before passage.

Today the Utah Legislature concludes the fifth week of its seven-week General Session. The pace is about to pick up incredibly. I’ve collected new data on how the Utah House and Utah Senate spend their time on the floor. I’ll use this new data to illustrate what’s coming in the next two weeks.

I include here only time spent on the floor of each chamber considering bills. I exclude time hearing the governor’s “State of the State” address, time in recesses or “saunters,” time on personal privilege motions, time hearing committee reports, and so on. I include time debating bills, voting on bills, and making motions directly relevant to a specific bill.

The first chart plots the number of bills heard each day in the Utah House and Utah Senate. If a bill was heard even briefly on a particular day, such as a 30-second motion to “circle” the bill (that is, a motion to delay consideration of a bill) it appears in the daily counts. There is a separate vertical line for each day, red for the Utah House and blue for the Utah Senate. The gaps are weekends and President’s Day. (Click any figure in this post to enlarge it.)

Bills considered per day

House rules require considering each bill only once before passage, but Senate rules require considering each bill twice on separate days. The Senate routinely waives this requirement in the end-of-session crunch. Thus, the Senate considers more bills per day than the House in the 4th, 5th, and 6th weeks (when the Senate follows its rules more faithfully), but the difference typically disappears in the 7th week (when the Senate does not).

The next chart shows the hours spent per day in the House and Senate considering bills. As bills progress from committees to the floor, each chamber schedules more time for floor debate, so there is an upward trend over the course of the session. The House typically spends more floor time overall considering bills than the Senate.

Hours spent considering bills on the floor each day

The real question, though, is how much time is spent per bill. Turns out I have a chart for that too. For each day of the session, this chart shows how many minutes were devoted to the average bill heard on the floor. I include all floor time on each bill over the course of a day. If HB42 was discussed for 5 minutes, then circled, then discussed for 3 more minutes later that same day, then that is added up to 8 minutes for that bill that day.

Average floor minutes per bill

A few outlying spikes cause other bars to get compressed down so small that the chart is nearly unreadable. Here is the same data, but omitting outliers where the average exceeded 20 to make the trends more apparent:

Average floor minutes per bill (outliers removed)

Notice that the House consistently spends more time per bill than the Senate. Notice also that the Senate is more consistent over the course of the session than the House, which visibly spends less time per bill in the final week than earlier in the session.

Both chambers have spent less time per bill as the bill load has increased. In this next chart, the red line shows the total number of bills passed per year, growing 23% from 436 in 2009 to 535 in 2018. The blue line shows that the median floor time per enacted bill (in seconds) has steadily declined from over 800 seconds per bill to around 700 seconds. It’s unavoidable: As the Legislature passes more bills, it will have less time to consider each one carefully.

More bill enactments means less time (in seconds) per bill

I’ve been speaking of the average or median time spent per bill. To be sure, floor consideration is not evenly distributed among bills. The next figure plots a separate vertical line for each bill that was considered in 2018, with enacted bills in the left panel and other bills in the right panel. Each vertical line represents the total number of floor minutes (across both chambers) spent on a particular bill. I’ve sorted the bills by how much total time was spent on them. You can see that a few outlying bills received more than 40 minutes of total consideration, time, but the large majority of enacted bills received less than 20 minutes of floor consideration.

Total floor time per bill (2018)

Let’s pick this figure apart a bit more. The median bill enacted in 2018 received a total of 11.8 minutes of floor consideration. Those 11.8 minutes include time spent (in either chamber) on motions relevant to the bill, debates about the bill, and voting on the bill.

Let’s deduct from this 11.8 the time spent voting. In my experience, it takes roughly 2 minutes for a typical House vote and roughly 3 minutes for a typical Senate vote. These are ballpark guesses. Subtracting roughly 5 minutes from 11.8, I estimate that the median enacted bill received something like 7 minutes of actual debate. Total. Across both chambers.

Louder for the folks in back: The median bill enacted in 2018 received a total of only seven minutes of House and Senate floor debate before passage. We need to have a serious discussion about whether that’s enough, and about whether the Legislature has sufficient resources to vet bills properly before passing them.

Here’s the same data for each session going back to 2009. This time I’ve marked enacted bills in blue and other bills in red, then overlaid them into the same figure. I showed earlier that the median floor time per enacted bill has shrunk over the past decade. You can see that trend here, too: A lot more bills crossed the 50-minute mark in the first 5 years shown than in the latter 5 years.

Total floor time per bill, by year

I’ve been writing about the Legislature’s workload and vetting for nearly a decade now, though this is the first time I’ve pulled together data on how floor time gets used. There’s more I’d like to collect. I’d like to collect data on each bill’s impact (it’s fiscal impact, the amount of Utah Code affected, and so on) to see whether there is a relationship between impact and floor consideration. Trends reported here would be less concerning if the bills receiving scant consideration tend to make only slight updates to state law.

I’d also like to collect data on committee consideration. The Legislature schedules much more time for committee hearings than for floor debate. That will be harder to compile than the floor time data, unfortunately.

It may take me another decade to collect the additional data I’d like. It turns out that this sort of hobby research isn’t really part of my job description at BYU. But hey, if the Legislature wants to pass HB183 this year to fund an outside study of its workload and capacity, well, I know a guy who might be interested.

Posted in Everything | Tagged , , , , | Comments Off on New data show how little time the Legislature spends per bill

Cage match: LaVar Christensen vs Dan McCay

Rep. LaVar Christensen and Rep. Dan McCay are squaring off to fill a vacant seat in the Utah Senate. These two have served a long time together in the Utah House–long enough to cast a lot of opposing votes. So let’s compare their voting records.

Summary statistics

Let’s start with some statistics summarizing each legislator’s service, drawn from my statistical tables about the Utah Legislature. The statistics below include only service from 2011 on. For statistics where I don’t have a long term average handy, I have included the 2018 value.

Christensen McCay
General Sessions in House 2003-07, 2011-18 2012-18
Bills sponsored 7.6/year 9.0/year
Bills passed 4.3/year 4.0/year
Senate bills floor sponsored 1.1/year 10.1/year
Missed votes 5.4% of 5,441 8.6% of 4,790
Nay votes 4.5% of 5,441 14.1% of 4,790
Party support 97.7% in 2018 86.1% in 2018
Party support on party-line votes 84.1% in 2018 94.0% in 2018

Christensen and McCay sponsor and pass a similar number of their own bills, but there is a huge gap on floor sponsorship of Senators’ bills. Senators carefully choose Representatives to carry their bills through the House, so the difference in floor sponsorship rates suggest different reputations among Senators. In fact, this difference in floor sponsorship strikes me as the most telling difference in this table; one of these legislators has a track record over the past several sessions of working closely across chambers, and the other does not.

For missed votes, the chamber average each year is generally around 5-7%. Christensen tends to fall right within this average, with McCay slightly above it, but neither is particularly far from the average.

McCay is famous for his “nay” voting record, and particularly for voting “nay” on commemorative (i.e. non-policy) resolutions. He casts “nay” votes at more than triple the rate Christensen does.

Party support scores measure how often these legislators vote with the majority of their own party. In 2018, Rep. Christensen voted with the majority of House Republicans 97.7% of the time, versus 86.1% for Rep. McCay. The more telling measure, however, is how often these legislators vote with their party when there is a “party-line” vote–that is, a vote wherein a majority of Republicans votes opposite a majority of Democrats. On party-line votes the scores flip, with Christensen voting with his party 84.1% of the time versus McCay’s 94.0%.

Specific votes

Over the past 7 General Sessions, Christensen and McCay have participated together in 4,136 votes, disagreeing 627 times (15%). (I exclude votes where one or both was absent.) That means 85% of the time it wouldn’t matter whether Christensen or McCay is the one casting a vote–you would get the same result either way.

If we narrow it to close votes–that is, where there were no more than 55 of 75 Representatives on the same side of the vote–then Christensen and McCay have participated in 555 votes, disagreeing on a phenomenal 272 of them (49%). That’s enough to make it hard to tell from the record whether they even belong to the same party.

Of these 272 disagreements on close votes, 3 were 38-37 decisions. That is, 3 were decided by a single vote, and either Christensen or McCay could have changed the outcome by switching his vote.

These 3 pivotal votes appear at the top of the table below, which lists every time Christensen and McCay cast opposing votes when the overall “ayes” vs “nays” margin was 9 or less. (The table is sorted so that votes with the closest “ayes” vs “nays” margins appear at the top.)

Bill Vote (ayes-nays-absent) Christensen McCay
HJR018 (2016) House/ passed 3rd reading (38-37-0) no yes
HB0176 (2017) House/ passed 3rd reading (38-37-0) yes no
HB0348S02 (2017) House/ passed 3rd reading (38-37-0) yes no
HB0424 (2012) House/ failed (36-37-2) yes no
HCR006S01 (2015) House/ passed 3rd reading (38-36-1) no yes
HB0428S01 (2017) House/ passed 3rd reading (38-36-1) yes no
HB0278S01 (2013) House/ failed (37-35-3) yes no
HB0358 (2014) House/ failed (36-38-1) no yes
SB0097S03 (2014) House/ failed (35-37-3) yes no
HB0078 (2017) House/ failed (34-36-5) no yes
HB0406S01 (2015) House/ passed 3rd reading (39-36-0) yes no
HB0298S03 (2012) House/ floor amendment (38-35-2) yes no
HB0381 (2013) House/ passed 3rd reading (38-35-2) yes no
HB0210S01 (2012) House/ failed (36-39-0) no yes
HB0068 (2018) House/ failed (34-37-4) no yes
HB0164S02 (2018) House/ failed (33-36-6) yes no
SB0078 (2013) House/ passed 3rd reading (38-34-3) yes no
HB0291S01 (2017) House/ floor amendment failed (35-39-1) no yes
HJR014 (2015) House/ failed (34-38-3) no yes
SB0251S03 (2016) House/ floor amendment # 1 (40-35-0) no yes
HB0395S05 (2017) House/ passed 3rd reading (39-34-2) yes no
HB0414 (2012) House/ passed 3rd reading (38-33-4) yes no
HB0091S01 (2014) House/ failed (35-40-0) yes no
HJR015S02 (2015) House/ failed (32-37-6) yes no
HB0376S01 (2015) House/ passed 3rd reading (40-34-1) no yes
HB0069S01 (2015) House/ passed 3rd reading (39-33-3) yes no
HB0409 (2014) House/ passed 3rd reading (39-33-3) yes no
HB0011S02 (2016) House/ failed (32-38-5) yes no
HB0326 (2017) House/ failed (32-38-5) no yes
SB0041S05 (2012) House/ passed 3rd reading (39-32-4) yes no
SB0150 (2015) House/ passed 3rd reading (39-32-4) yes no
HB0260S01 (2018) House/ passed 3rd reading (39-32-4) yes no
SB0115S04 (2016) House/ failed (33-40-2) no yes
SB0033S01 (2015) House/ failed (32-39-4) yes no
HB0274 (2013) House/ passed 3rd reading (41-33-1) yes no
SB0134S03 (2015) House/ passed 3rd reading (41-33-1) yes no
HB0220S01 (2016) House/ passed 3rd reading (41-33-1) yes no
HJR008 (2016) House/ passed 3rd reading (41-33-1) no yes
HB0254S01 (2018) House/ passed 3rd reading (41-33-1) no yes
HB0088S02 (2013) House/ passed 3rd reading (40-32-3) yes no
SB0136S06 (2018) House Conference Committee – Final Passage (40-32-3) yes no
HB0246 (2013) House/ concurs with Senate amendment (39-31-5) yes no
SB0078 (2015) House/ passed 3rd reading (39-31-5) no yes
HB0163S02 (2018) House/ passed 3rd reading (39-31-5) no yes
SB0129 (2018) House/ passed 3rd reading (38-30-7) no yes
SB0099 (2018) House/ failed (31-39-5) yes no
HB0079S01 (2015) House/ passed 3rd reading (41-32-2) yes no
HB0246 (2013) House/ passed 3rd reading (41-32-2) yes no
HB0447S02 (2015) House/ passed 3rd reading (40-31-4) yes no
HB0332S01 (2016) House/ passed 3rd reading (40-31-4) yes no
SB0284S01 (2013) House/ passed 3rd reading (38-29-8) no yes
HJR008S01 (2014) House/ failed (32-41-2) no yes
HB0394S02 (2015) House/ failed (32-41-2) no yes
SB0149S02 (2016) House/ failed (32-41-2) yes no
HB0360S03 (2017) House/ failed (32-41-2) yes no
SB0097S03 (2014) House/ floor amendment failed (31-40-4) no yes
HB0187S01 (2016) House/ failed (31-40-4) yes no
HB0284 (2018) House/ failed (30-39-6) no yes
HB0307 (2013) House/ failed (29-38-8) no yes

Opposing votes matter more on close votes than on lopsided ones, so I’ll cut the table off there. If you really want to see comparisons where the overall margin was greater than 9, email me I guess.

Posted in Everything | Tagged , | Comments Off on Cage match: LaVar Christensen vs Dan McCay

Analysis of voting patterns in the 2018 Utah Legislature

The Utah Legislature concluded its 45-day General Session last Thursday at midnight. Once again, I’ve scraped the voting records to produce this statistical summary.

Update (3/14/2018): The Utah Legislature’s website initially showed HB457 as having passed. Their records were corrected a few days after the session ended (and after I posted this) to show that it did not receive its final vote before the constitutional deadline of midnight on March 8. Thus, the total number of passed bills is 533, not 534. I have not made changes to this post, but I have corrected the underlying data tables that are available at my other website.

Number of bills considered and enacted

There was a great deal of hype before and during the session about legislators preparing a record number of bills, and one unnamed legislator planning as many as 83.

It appears, though, that legislators reached the natural limit for bill introductions and passages long before 2018. Legislators may have filed behind-the-scenes preparatory work on a record 1,359 bills this year, but in the end only 817 bills were actually introduced–in line with bill introductions from the past several years. (Throughout this post, click on charts to enlarge them.)

Let’s be clear: 817 bills is a tremendous number to consider in a 45-day session, and introducing so many bills places tremendous pressure on vetting. Considering 817 bills–and passing 534 of them, one shy of last year’s record 535–requires forcing many of them through in an awful hurry. The average enacted bill aged only 16.2 days before its first floor vote, and 26.9 days before its final vote.

Running up the score? The most active bill sponsors in 2018

Rep. Steve Eliason introduced the most bills this year (25), passing 16, followed by Sen. Todd Weiler (21, passing 17); Sen. Wayne Harper (19, passing 16); and Sen. Daniel Thatcher (18, passing 16). (I exclude floor sponsorship and cosponsorship here; this is primary sponsorship only.)

Several legislators tied for 5th place with 17 bill introductions: Sen. Curt Bramble (passed 16/17), Sen. Karen Mayne (passed 15/17), Rep. Mike McKell (passed 11/17), and Rep. Stephen Handy (passed 10/17).

Together, these 8 legislators introduced 151 bills. That is, 8% of legislators introduced 18% of bills.

No legislator introduced zero bills except for Rep. Lynn Hemingway, who missed the entire session due to his wife’s grave illness; Rep. Jon Stanard, who resigned mid-session; and Rep. Travis Seegmiller, who filled Standard’s seat halfway through the session. But three legislators introduced only 1 bill: Rep. Susan Duckworth (passing 0), Rep. Susan Hollins (passing 1), and Rep. Keith Grover (passing 1).

Find complete bill sponsorship statistics for each legislator here.

Late bill introductions

With only 7 weeks in the General Session, legislators needs to have their bills ready on day 1 in order to have sufficient vetting time. They did better this year than in a long time, with 392 bills ready to introduce in week 1. That’s the most introductions we’ve seen since the Legislature changed the General Session’s start date in 2009. (Oddly enough, moving the start date *back* a week in 2009 caused bills to be ready *later* in the session rather than earlier, something I’ve written about before.)

At the same time, we still saw an awful lot of bill introductions in the session’s second half: 146 bills were introduced in the final three weeks, much more than last year’s 108, though still an improvement over 2011-2016. Simply put, if we’re going to have a 45-day session, then more bills need to be ready to go on day 1.

Bipartisan consensus remains the rule

As in the past, bipartisan consensus remains the rule for floor votes in the Utah Legislature. In 2018, the average Utah House floor vote witnessed 94% of the body voting together–a figure that rises to 97% in the Senate. These high rates have been common for many years in the Legislature, as the figure shows.

The flip side is that party-line votes, in which a majority of Republicans vote against a majority of Democrats, are rare. Only 1 in 10 House votes were decided along party lines in 2018, dropping to 1 in 20 in the Utah Senate. These figures were even lower than usual, but not by much–we’ve always been pretty close to the floor by this metric.

It’s not kumbaya for everybody, of course. When party-line votes do arise, the most loyal partisans were Sen. Ralph Okerlund and Rep. Jon Stanard (who voted with their party 100% of the time on party-line votes), Rep. Keith Grover (98.4%), Rep. Mike Noel (96.%), Sen. Don Ipson (95.9%), and Rep. Rebecca Chavez-Houck (95.8%).

The least loyal partisans were Sen. Brian Zehnder (43.8%), Rep. Becky Edwards (59.4%), Rep. Bruce Cutler (62.5%), and Rep. Craig Hall (62.9%). Find full rankings here. (Curiously, Sen. Zehnder’s sub-50 score means that on party-line votes, he voted with Democrats more than with his own party.)

For the curious, the closest six votes in the 2018 House were on HB197 (36-34-4), HB68 (34-37-3), HB164 (33-36-5), HB313 (38-33-4), HB323 (38-33-4), and HB471 (34-39-2). The closest six votes in the 2018 Senate were on HB116 (11-11-7), HB396 (12-12-5), HJR11 (12-12-5), SB52 (13-13-3), SB91 (12-12-5), HB309 (11-12-6). Find a longer list of close votes from each chamber here.

A fair chance for Democrats

Democrats hold just under one-fifth of the seats in each chamber of the Utah Legislature. That means Republicans control each chamber’s floor debates, Rules Committee, and standing committee chairmanships, any of which could be used to prevent Democratic-sponsored bills from even coming to a vote. If the Democrats were a large enough minority to threaten Republicans, then we would surely see Republicans using these majority powers to stifle Democrats, much as we see in Congress.

We don’t, though. Utah Republicans feel no threat from Utah Democrats, so they give most Democratic bills a fair shake. In 2018, 66% of Republican-sponsored bills passed, compared to 57% of Democratic bills. We’ve seen smaller gaps–6% in 2016–but this 9% gap is about as small as it gets. House Speaker Hughes and Senate President Niederhauser are loyal Republicans, yet they do not use their power to stifle Democrats. (Of course, Democrats’ high success rates, along with the low party-line voting rates reported above, also reflect a Democratic strategy of “go along to get along,” where they save their political capitol for their highest priority issues.)

Legislators abdicate tough choices to leadership

Utah legislators are “Utah nice.” That is, many prefer not to publicly vote against their colleagues’ bills. In 2018, only 1% of Utah Senate floor votes had a negative outcome; it was 3% in the Utah House. So if 817 bills were introduced but only 534 passed, what happened to the other 283 bills? Funny you should ask……

When bills are introduced on time so that they can receive their committee hearing early in the session, legislative rules require all bills to receive floor consideration in the order they emerge from committee. But at the end of the session, that goes away–both chambers wipe their calendars and send all remaining bills to the leadership-aligned Rules Committee for prioritization. From that point forward, bills are heard in the order prioritized by each chamber’s leadership. And since 301 of the 534 bills (56%) that passed in 2018 received didn’t pass until the last week, nearly all bills that pass had to survive this sifting process.

As such, we might conclude that legislators have collectively abdicated to legislative leadership their responsibility to vote “no.” True, committees take an active role in amending bad bills into something more palatable, but for the most part, bills fail not because they were voted down, but rather because leadership did not prioritize them. I leave it to the reader to decide whether rank-and-file legislators should take a more active role deciding which bills pass. In any event, this figure shows that low failure rates have been common for a long time:

Legislators who vote “nay”

Some legislators vote “nay” more than others, of course. This year, the most “nay” votes were cast by Rep. Marc Roberts (13.9%), Rep. Dan McCay (13.6%), Rep. Brian Greene (13.2%), and mid-session newcomer Rep. Travis Seegmiller (13.1%).

“Nay” votes are less common in the Senate; the highest “nay” voter there, Sen. Margaret Dayton, voted only 9.6% “nay,” which would rank only #11 in the House. After her, it’s a long way down to Sen. Jim Dabakis (6.8% “nay”), #2 in the Senate but tied for #32 if he served in the House.

The other end of the rankings are dominated by Senators, though. The #1 ranking for least “nay” votes goes to Sen. Ralph Okerlund and Senate President Wayne Niederhauser, tying at 0.2% “nay.” They are followed by Sen. Daniel Thatcher (0.4%), Sen. Stuart Adams (1.1%), and Sen. Jerry Stevenson (1.2%).

The fewest “nay” votes in the House were cast by Speaker Greg Hughes, with 1.5% “nay.”

Absentee rates

Each chamber had a legislator miss a large number of votes due to unfortunate circumstances. In the House, Rep. Lynn Hemingway missed the entire session due to a family medical issue; in the Senate, Sen. Ralph Okerlund had health issues of his own. I omit them from this discussion.

The best attendance awards go to Rep. Kay Christofferson (missed only 0.4% of votes), Rep. Susan Pulsipher (0.7%), Rep. Marie Poulson and Rep. Jefferson Moss (tied at 0.8%), and Rep. Mike Winder (1.0%). Given the insanely fast pace of floor voting in the session’s final days, even using the restroom can mean missing a few votes, so these attendance rates are unquestionably impressive.

At the other extreme we find the usual suspects: Legislative leaders and budget chairs, who miss lots of floor time while negotiating compromises and finalizing the budget. Other than Rep. Hemingway and Sen. Okerlund, the most floor votes were missed by Senate budget chair Jerry Stevenson (32.6%), House Speaker Greg Hughes (30.5%), Senate President Wayne Niederhauser (29.7%), House vice budget chair Mike Schultz (26.4%), House majority whip Francis Gibson (23.0%), and House budget chair Brad Last (22.6%). Find absentee rates for all legislators here.

Ideology scores

And now we come to ideology scores. I calculate these each year using the W-NOMINATE algorithm developed by Keith Poole and Hal Rosenthal in their study of Congress. You can read about their efforts here. It seems that these get misused in campaign mailers and occasionally by the media when I release them each year, so let me stress some caveats.

(1) Scores have no scale. I constrain them to fit between -100 (relatively more liberal) and +100 (relatively more conservative), but there is no such thing as a “liberal” or “conservative” score. The ideology that corresponds to a -100 or +100 score will depend entirely on the mix of people serving in the Legislature at a given time.

(2) Scores are calculated separately for each chamber and each year. Thus, you cannot directly compare a House score to a Senate score, or a 2017 score to a 2018 score.

(3) You should not brag about having the “most conservative” or “most liberal” voting record in your campaign materials, especially if your intent is to imply that “most conservative” means “most faithful to the Republican platform.” A score of +100 could mean “most extreme” or “most beyond the mark” just as easily as it means “most faithful to conservative principles,” depending entirely on what the range of people in the Legislature was this year. You need to look at an individual’s actual voting record to decide whether it was “most faithful to party principles” as opposed to “most extremely beyond party principles.” Refer again to point #1.

(4) Zero does not mean “moderate,” and scores less than zero are not “liberal.” Scores have only relative meaning, not absolute meaning. Again, see point #1.

To be clear, you are committing statistical malpractice if you use these statistics in campaign materials or news reporting without putting them into context with caveats like those above, and you’d better believe I’m going to call you out for it.

Okay. Have I belabored this point enough? Here we go. The plot below depicts the distribution of ideology scores in the 2018 Utah House and Senate.

Two things to notice:

(1) Both chambers have a “bridge” legislator between the parties. In the House, that’s Democrat Susan Duckworth; in the Senate, it’s Republican Brian Zehnder, who, as noted earlier, voted with Democrats more than Republicans on party-line votes.

(2) In both chambers, there is much more ideological spread among Republicans than among Democrats. Remember, these scores can only detect differences that were reflected in actual floor votes. Ask yourself: In a supermajority Republican legislature, how much of the floor action is going to be over bills where the liberal and moderate wings of the Democratic party are at loggerheads and trying to find a compromise? None, right? Now ask yourself how many are going to split the moderate and conservative wings of the Republican party. Simply put, supermajority Republican control brings out factional disputes among Republicans while obscuring factional disputes among Democrats. Thus, year after year, these scores reveal much more ideological daylight between Republicans than between Democrats.

Within the House, the more extreme wing of the Republican party is anchored by Dan McCay (94), Brian Greene (90), Marc Roberts (90), Norm Thurston (76), Adam Robertson (76), and Karianne Lisonbee (74). The more moderate wing is anchored by Becky Edwards (-26), Craig Hall (-18), Eric Hutchings (-16), Bruce Cutler (-10), and Ray Ward (-10).

Within the Senate, the more extreme wing is anchored by Margaret Dayton (98), David Hinkins (90), Allen Christensen (90), and Deidre Henderson (88), with the more moderate wing anchored by Brian Zehnder (-55, with a score closer to Democrats than to the nearest Republican), Todd Weiler (2), Lyle Hillyard (5), and Gregg Buxton (15).

Find W-NOMINATE scores for all legislators (but only if you remember the caveats!!) here: House and Senate.

One more word about Sen. Zehnder. Though we cannot directly compare scores from different years, we can make reasonable statements about the relative position of a legislator compared to his peers in one year versus the next. In the figure below, the left panel shows the distribution of scores for the 2017 Senate, while the right panel shows the distribution for the 2018 Senate. At left, Sen. Brian Shiozawa was the most left-leaning Republican in 2017, with a score -23; when he resigned last December, he was replaced by Sen. Brian Zehnder, this year’s most left-leaning Republican, with a score of -55. Democrats appear to be targeting Zehnder’s seat; Kathie Allen and Kathleen Riebe have already announced candidacies, and Wayne Holland is rumored to be considering a run. I’ll let the reader work out the implications–both how these rumored runs may have impacted Sen. Zehnder’s voting behavior, and also how this voting record may impact these challenges.

Concluding remarks

If you read all the way to this point, you must be getting paid for it. Remember, the source statistics for everything in this post can be found here. If you plan to publicize a statistic, please double check it from that source, since I’ve probably made a typo or two in this post. I maintain this blog as a hobby, so I’m not exactly getting paid to go back and double check this post–but I use the underlying stats in my actual scholarly research, so I’m more careful with the source.

The mid-session change from Rep. Stanard to Rep. Seegmiller caused some weird glitches in my code, as did the full-session absence of Rep. Lynn Hemingway. Neither of those things has happened in the 12 sessions that I’ve been doing this. I think I caught and fixed all the errors, but please tell me if you spot something I missed.

And hey, let me end with this: Pre-order my book about Utah politics from Amazon, forthcoming this summer from the University of Nebraska Press. It will be the best book you ever read, obviously.

Posted in Everything | Tagged , , , , , , , | 1 Comment

What Rep. Watkins teaches us about party and ideology

Rank-and-file legislators know which side their bread is buttered on.

Political scientists have made a parlor game of calculating ideology scores for elected officials based on their voting records. The gold standard for the US Congress is the DW-NOMINATE algorithm; you can find those scores back to the founding at Keith Poole’s voteview.com. Shor and McCarty have calculated similar scores for legislatures from all 50 states.

Just for fun, I apply the W-NOMINATE algorithm each year to the Utah Legislature to infer each legislator’s relative ideology. Let me give some caveats first.

Caveat #1: These scores have no absolute meaning; there is no score that means “conservative” or “liberal,” for example. But they do have relative meaning; if Legislator A has a higher score than Legislator B, then we can say that Legislator A has a voting record to the right of Legislator B. But to be clear, the highest score is not necessarily the “most conservative” or “most Republican” score, since somebody from the lunatic fringe could get a very extremely far-right or far-left score.

Caveat #2: Sometimes I hear of legislators boasting to the party faithful that I gave them the “most conservative” score. That is a misuse of these data. See Caveat #1.

Caveat #3: I calculate these scores separately each year. Because these scores have no absolute scale, that means you can’t compare a score from one year to a score for another. Nor can you compare a House score to a Senate score–if you want to do that, use the Shor-McCarty scores instead.

Having said all that, take a look at the distribution of scores for the 2017 Utah House. (You can find the Senate, or other years of the House, here.) I constrain scores to fall between -100 (more liberal) and +100 (more conservative). Again, the midpoint doesn’t mean “moderate,” it means “the middle of the House,” which will be a pretty conservative position. Here’s the 2017 Utah House, with a “D” marking each Democrat’s score and an “R” marking each Republican’s:

Ideology - Watkins 2017

See the one that I highlighted? That’s Rep. Christine Watkins. She served 4 years as a Democrat from 2009-2012. After losing reelection, she switched parties and rejoined the Legislature this year (after a 4-year absence), this time as a Republican. Now take a look at the distribution of ideology scores for the 2012 Utah House, Rep. Watkins’s last year as a Democrat. I’ve highlighted her again.

Ideology - Watkins 2012

Now, I realize I just said you can’t compare a score from one year to a score from another. But we’re not doing that. We’re comparing her relative position in one year to her relative position in another. In 2012, she was within the Democratic mainstream, though on the conservative side. In 2017, she was within the Republican mainstream, though on the liberal side.

Let’s look at this another way. If you look up Rep. Watkins (2017) in my “who votes together” tool, you’ll find that the person she voted with most often in 2017 was  (drumroll please) Speaker Greg Hughes. They voted together 95.3% of the time. Meanwhile, out of her 74 House colleagues, she voted less often with Democratic minority leader Brian King than with all but 6 of colleagues (also Democrats).

Now back up to 2012. If you look up Rep. Watkins (2012) in my “who votes together” tool, you’ll find that she voted with Brian King 93.5% of the time that year. (He wasn’t minority leader that year, but let’s use him for comparison anyway.) She voted with him more often than anybody else that year, except for 2 other Democrats who barely edged him out. What a difference 4 years make.

To be clear, this is not a post about Rep. Watkins. It’s a post about the dance of ideology and partisanship. It’s possible that she experienced a profound ideological conversion during her absence from the Legislature. It’s more likely that we’re seeing the influence of party leadership on her floor voting. I don’t mean to imply that party leaders told her how to vote, but rather that rank-and-file legislators know which side their bread is buttered on.

You can find ideology scores for all legislators here. And you can find my “who votes together” tool here.

Posted in Everything | Tagged , , , | Comments Off on What Rep. Watkins teaches us about party and ideology

Who voted “nay” the most in the 2017 Legislature?

The House had its lowest failure rate in half a decade.

As I posted earlier, the Utah Legislature is almost a bipartisan lovefest. Legislators just don’t like voting “nay.” In general, if something gets to the floor, it’s going to pass. In 2017, only 2% of Utah House floor votes had a negative outcome. It was even lower in the Utah Senate: 1%. Though the Senate has been at 1% for years, the House had its lowest failure rate in half a decade. Here’s the trend:

Failure rates, 2007-2017

Of course, some legislators vote “nay” more than others. Surprising nobody, Speaker Greg Hughes and President Wayne Niederhauser seldom vote nay; after all, if they don’t like a bill, they can just kill it. More generally, legislative leaders often have very agreeable voting records. And surprising nobody, Democrats often vote nay more than Republicans. That’s to be expected when they’re in the minority. Where it gets fun is looking at the Republicans with high “nay” rates.

The lowest “nay” rates in the House: Representatives Hughes (1%), Noel (3%), Wilson (3%), Christensen (3%), and Handy (3%). The highest “nay” rates: Representatives Kwan (14%), Briscoe (14%), Hollins (15%), Chavez-Houck (15%), and Romero (16%). Among Republicans, the highest “nay” rates were Representatives Greene (12%), Roberts (13%), and McCay (13%). In years past, Rep. McCay voted “nay” more than anybody in the chamber, including the Democrats; this is the second year in a row where he’s fallen behind at least a few Democrats.

The lowest “nay” rates in the Senate: Senators Jerry Stevenson (0%), Hemmert (1%), Buxton (1%), Niederhauser (1%), and Okerlund (1%). The highest (and this time there are Republicans in the top 5): Senators Hinkins (5%), Iwamoto (5%), Escamilla (6%), Dayton (7%), and Dabakis (9%). (Senators Hinkins and Dayton are the Republicans.)

You can find “nay” rates for any legislator here.

Next up: How Representative Watkins’s change from Democrat to Republican changed her voting record. (Hint: It’s even more fascinating than you’re probably guessing.)

Posted in Everything | Tagged , , | Comments Off on Who voted “nay” the most in the 2017 Legislature?

Who missed the most votes in the 2017 Legislature?

Absenteeism declined  in 2017.

My previous two posts had some good news about the 2017 Legislature: Vetting time improved a little, giving the public more time to see most bills, and bipartisanship remained the order of the day. But now it’s time for a more dubious distinction: Calling out those who missed the most votes.

First, the usual caveats. In a 45-day session with hundreds of bills to consider, it is inevitable that legislators will need to leave the floor at times. There are literally times near the end of the General Session when leaving the floor for an urgent restroom break can cause a legislator to miss several votes. So all legislators can be excused for missing a handful of votes.

Second, if a legislator experiences a health issue or attends a funeral for a day, that can also cause several missed votes–especially if these issues arise toward the end of the session.

Third, legislative leaders and those with responsibility over the process (i.e. the Rules Chair) or the budget are expected to miss many votes toward the end of each session.

Of course, these aren’t the only reason that legislators miss votes. In the 2017 Legislature, 10% of Senators were absent from the average floor vote. This is actually an improvement over recent years; we haven’t seen better since 2007. The House also improved, pushing average absenteeism down to only 5%. So absenteeism declined  in 2017. Here’s the trend for each chamber:

Absenteeism rates, 2007-2017

In the House, the almost perfect attendance awards go to Representatives Hall, Kennedy, and Westwood, each of whom missed only 0.3% of votes. At the other end are Representatives Ray (13.8%), Gibson (16.0%), Noel (16.6%), Wilson (20.6%), and finally Hughes (32.5%). Other than Paul Ray, these other 5 are all in chamber leadership.

In the Senate, former Representative and freshman Senator Gregg Buxton had the best attendance, missing only 2.7% of votes, followed by Senators Don Ipson (3.4%) and Jani Iwamoto (3.7%). I’ve always found it curious that even the best-attending Senators miss so many more votes than the best-attending Representatives. At the other end are Senators Hillyard (15.0%), Anderegg (17.0%), Hemmert (18.7%), Niederhauser (25.7%), and Stevenson (29.7%). Only 2 of these 5 are in leadership–Niederhause and Stevenson. The next 2 (Anderegg and Hemmert) are new to the Senate this year.

You can find absentee rates for all legislators who served in 2017 here.

Next up: Who voted “nay” the most?

Posted in Everything | Tagged , , , | Comments Off on Who missed the most votes in the 2017 Legislature?

Utah lawmakers loved to agree in the 2017 Legislature

Most bills that make it to a vote pass comfortably with bipartisan support.

The partisan rancor that pervades national politics seldom reaches the Utah Legislature. Simply put, Republicans control such an overwhelming supermajority of seats that they have no need to fear Democrats. And since they don’t fear them, they are more open to their ideas than a newcomer to Utah politics might expect.

Many metrics reveal this pattern. We can start with each party’s batting average. Once again, Democrats passed over half their bills. They’ve done worse under more some speakers than others (take a look at the pre-Lockhart and pre-Hughes era). But their 53% bill passage rate for 2017 is downright respectable for a party that holds less than 1 in 5 seats in each chamber.

Party batting averages, 2007-2017

Let’s look at floor votes. Party line votes–that is, votes where the majority of legislative Republicans votes against the majority of legislative Democrats–are almost unheard of in the Utah Legislature. The next figure shows that only 13% of House votes and 6% of Senate votes were decided along partisan lines.

Party line voting, 2007-2017

Instead, most bills that make it to a vote pass comfortably with bipartisan support. In 2017, 93% of legislators voted together on the average House vote; it rose to 97% in the Utah Senate. That 97% figure actually marks a record in my 11-year data series. Still, it’s not far from the typical range, as you can see:

Average size of floor vote coalitions, 2007-2017

There were some doozies, of course. Here are the 12 votes decided by 8 Representatives or fewer (i.e. by 10% or less of the 75-member body) in the Utah House:

Bill Vote type Ayes-Nays-Absent Vote margin
HB0029S04 House/ failed 37-38-0 1
HB0176 House/ passed 3rd reading 38-37-0 1
HB0348S02 House/ passed 3rd reading 38-37-0 1
HB0078 House/ failed 34-36-5 2
HB0408 House/ passed 3rd reading 38-36-1 2
HB0428S01 House/ passed 3rd reading 38-36-1 2
HB0178 House/ floor amendment # 1 38-34-3 4
HB0291S01 House/ floor amendment failed 35-39-1 4
HB0395S05 House/ passed 3rd reading 39-34-2 5
HB0326 House/ failed 32-38-5 6
SB0114S06 House/ substituted from # 4 to # 6 40-33-2 7
HB0156S01 House/ passed 3rd reading 40-32-3 8

And here are the 6 votes decided by 3 Senators or fewer (i.e. by 10% or less of the 29-member body) in the Utah Senate:

Bill Vote type Ayes-Nays-Absent Vote margin
SB0029 Senate/ failed 14-14-1 0
HB0099S01 Senate/ passed 2nd & 3rd readings/ suspension 15-14-0 1
SB0088 Senate/ failed 12-13-4 1
SB0115S01 Senate/ failed 13-15-1 2
HB0149S02 Senate/ passed 3rd reading 15-12-2 3
SB0115S01 Senate/ failed 13-16-0 3

Here is a longer list of close votes from each chamber, and here are more tables and charts with floor voting patterns in the Utah Legislature.

Up next: Who missed the most votes?

Posted in Everything | Tagged , , , | Comments Off on Utah lawmakers loved to agree in the 2017 Legislature

The 2017 Utah Legislature passed a record number of bills but slightly improved vetting time

We’re not back to the good old days of 2007-2008, but legislators definitely did better this year at getting their bills out earlier

Back in January, I heard a lot of chatter that this would be one of the busiest sessions ever. Folks were saying that legislators had submitted some 1200 bill requests to drafting attorneys. Though it’s true that the 2017 Legislature passed a record number of bills–535, edging out the 528 from two years ago–there weren’t necessarily more bills on the docket. Legislators introduced only (only!) 810 bills, a decline from the 831 introduced two years ago. As you can see from this next chart, this isn’t much of a change either way.

Total bills introduced and passed, 2007-2017

Legislators continued their habit of rushing bills through. With only 45 days for the General Session, the most any bill will age between introduction and final passage is 45 days. Of course, few bills actually age that much. If legislators don’t get their bill requests to staff before the holidays, then the bills won’t be ready for introduction until halfway through the session. Of course, most bills don’t get nearly 45 days of vetting. In 2017, the average bill aged only 17.2 days from introduction to first floor vote. As fast as that sounds, it’s actually the longest interval since the 17.9 day average clear back in 2007.

After that first vote on day 17.2, the average bill aged only 11.2 more days before its final passage. Those are calendar days, not working days; taking out weekends, 11.2 calendar days are only 9.2 working days. It’s hard for bicameralism to work when the second chamber gets so little time to consider bills, but at least we’re doing a little better than we were in 2011 and 2012.

You can see this improvement in the next chart. In the chart, the public has the most time to react to bills when the orange bars are tall and the blue bars are short. This year, 371 bills were introduced during the first week; after last year’s 376, that’s the most since 2008. Only 108 bills were introduced in the session’s final three weeks, the least since 2010. We’re not back to the good old days of 2007-2008, but legislators definitely did better this year at getting their bills out earlier, allowing more time for public comment. There were exceptions, of course.

Bills introduced early or late, 2007-2017

Next up: Floor voting patterns. Once again, bipartisan consensus rules the day, with minimal party-line voting and only a few close votes. Read it now.

Posted in Everything | Tagged , , | Comments Off on The 2017 Utah Legislature passed a record number of bills but slightly improved vetting time

How corrupt is Utah?

On the whole, Utah has lower-than-average levels of political corruption.

At length, former Utah Attorney General John Swallow is finally scheduled to face a jury this week. No matter how Swallow’s story ends, however, it marks an aberration. Federal statistics show that Utah has had very low levels of corruption and political malfeasance over the past several years.

I’m leaning on annual reports from the Public Integrity Section at the Department of Justice, which tabulate the number of federal, state, or local officials (whether elected or appointed) convicted of malfeasance in office each year. Recognizing that more populous states tend to have more elected officials, I use convictions per capita rather than total convictions. Well, actually I’m using convictions per 1,000,000 residents, which makes for more workable numbers.

The result? From 2002 through 2015, Utah averaged 1.4 convictions each year per million residents. Only 5 states did better: Oregon (1.0 per year), New Hampshire (1.0), Minnesota (1.2), South Carolina (1.2), and Washington (1.2). Some did far worse: Louisiana (9.3 per year), South Dakota (8.0), North Dakota (7.3), and Kentucky (6.8).

If you’re like me, the chart below will be easier to understand than the preceding paragraph. Click it to make it bigger and easier to read. You’ll find Utah near the top, ranked 6th out of 50 states, only marginally behind the leaders.

Annual corruption cases by state

Here’s another, slightly more complicated way to look at it. The next chart shows the yearly total separately for each state. You’ll see that Utah stays steadily low, while some states have spikes up and down. Again, click the image to enlarge it. States are ordered alphabetically this time, so Utah comes on the last row:

Corruption by state-year

Caveat #1: The Justice Department’s statistics count only convictions won in federal court (whether against federal, state, or local officials from a state). They do not count convictions won in state court. Of note, Swallow is being tried in state court. For that to shuffle these rankings, though, you have to persuade me that the feds are systematically less likely to bring charges in federal court in Utah than elsewhere.

Caveat #2: These statistics count only the number of convictions, with no consideration of the qualitative impact of an official’s malfeasance. This caveat matters a lot more than the preceding one: If Swallow really did what he’s accused of doing, then the scale of his corruption goes far beyond most abuse of office.

In any event, after some turbulent weeks for political news, isn’t it nice to hear something good? On the whole, Utah has lower-than-average levels of political corruption.

Posted in Everything | Tagged , , , , | Comments Off on How corrupt is Utah?