23 June 2011

Complete VR statistics for June 2011, compared to 2007 (and 2008)

Voter Registration in the USA: 2011
as compared to 2007 (and the GE 2008)

I have completed a major study of voter registration in the USA. Well, „completed“ is not accurate, for the stats are constantly changing. Rather, I can now provide a very reliable snapshot of one point in history and compare it as best as possible to the same time frame four years ago and again to the GE in 2008. These are the RV statistics for the Union in June 2011. I will be updating this data quarterly and posting a new report each time. For that is the entire idea – to follow and note changes in the electorate in the time leading up to the GE 2012.

The entire data I have collected is in one EXCEL document, which you can see here. In column „A“, all pertinent data is hyperlinked to the site of the document where you can see and/or download the data yourself. Only in the case of OHIO and RHODE ISLAND did I receive part or all of the necessary data directly from their SOS offices, so those two states are not hyperlinked. Should you have difficulty in accessing the hyperlinks, I can gladly send a copy of the excel document per email. Just ask for it.

Hard facts:

Currently, there are 177,777,895 registered voters in the continental USA plus DC.
Of those 177,777,895, 105,450,117 (59.32%) registered voters come from the 29 states plus DC that provide VR statistics by partisan affiliation.
The other 72,327,778 (40.68%) registered voters come from 19 states that do not provide VR statistics by party affiliation, but rather, publish a VR total.*/**
The picture will always be somewhat incomplete as ND has no voter registration at all and MS provides no VR stats.

*Actually, Illinois provides partisan information, at a cost of 500 USD per full download and cites a most curious state law as to why it cannot or will not provide the data directly. However, at the same time, some counties in IL provide very specific VR data by party affiliation quite freely over the internet, like Kane County, IL, and one county even provides an extensive voter list, by name and affiliation, ala Ohio, for download.: Madison County, IL So, setting aside the hypocrisy and lack of openness about this, for now I must classify IL as a state that only provides VR information as a VR total, without party affiliation.

**I have sent emails to the other 18 states that do not officially publish VR by party affiliation to see if they do indeed have an archive with such data after all. Should anything come up, then I will alter this report accordingly.

Here is a map of the US, broken down by states that publish VR by party affilation vs. those states that don't:

Of the states that do VR by party affiliation, here is the national breakdown:

Nationally, the Democratic party has a +12.03% registration edge among registered voters in the 30 states that publish tallies according to party affiliation.
But this is in many ways meaningless as there are states with majority DEM registered voters (WV, OK, KY, LA), but with a GOP voting history. Nonetheless, studying VR by party affiliation can be VERY helpful when we compare the statistics to the past.

There is another factor that makes this all difficult to quantify perfectly: states have different ways of enrolling and de-enrolling voters. Some states have active and inactive voter lists, some do not. Some states require their citizens to re-register between the primary season and the GE, others do not. Some states automatically take you off the rolls if you do not vote in the primary. Some states have drive-by registration (NC), others have same day registration. It can also take some time to remove a voter from the rolls when that person moves or dies. In terms of uniformity and because of federalism, this part of it all is a mess, frankly.

But there are some comparisons that indeed can help: before a presidential election, voter registration tends to spike throughout the Union. This is normal, for many voters who are not so-called „base“ voters (stalwart partisan voters) will register quickly and make their choice for President. Indeed, some citizens vote only for president and leave the rest of the ballot blank. We can prove this by looking at voter statistics from the election and see that many more votes were cast for President in this or that state than were cast for Governor or Senator. It is rarely -if ever- the other way around. Conversely, these voters tend to fall off the rolls in the off-years (the 1st and 3rd years of a presidential term), but especially in the year before the primaries start up, right around this time. So, comparing 2011 to 2007 can tell us some important things.

  1. It can tell us if there has been growth or decline over the last four years in terms of enrollment.
  2. It can tell us if the partisan mix has radically altered over the last four years.
  3. And, in a beauty contest with the GE-2008, it can tell us how much partisan strength has remained since the GE, but this data is statistically inconclusive.

I am not saying in any way that VR is a predictor of elections. Indeed, WV, KY, LA and OK prove that partisan registration does not necessarily translate to votes at the ballot box for one party or another. In the cases of WV, KY, LA and OK, we are with great certainty talking about conservative Democrats who are more in line with the Republicans on a number of issues but have maintained their former party status. However, VR studies in conjunction with knowing the electoral history of a state can give us an excellent idea where that state stands and if battleground tendencies are present. Here are examples:

Category I: absolute „Locks“.

With a massive GOP +45.38% registration edge, Wyoming is the most GOP friendly state in the Union. It has also widened its partisan margin considerably since 2006 (similar to CA, see below), by exactly +9.00%.

Wyoming's electoral history tells a similar story: It has gone for the Republican candidate for President for 14 of the last 15 election cycles and with the exception of 1992, with double digit margins. It was John McCain's strongest state in 2008 and unseated UT as the most republican state in the Union, based on „performance“ on that night (McCain +32.24%). Wyoming's congressional delegation is entirely Republican. The Governor and Secretary of State are Republican. The Wyoming House of Representatives (50 R - 10 D) and Wyoming Senate (26 R – 4 D) are overwhelmingly republican. Put bluntly, there are not enough Democrats in this state to even make a dent in the vote, except in a possible three man race (see 1992: Bush 41 - Clinton – Perot).

A mirror image of WY, DC is an absolute Democratic lock, with a massive DEM +69.19% registration edge. It, like WY, has widened its partisan margin over 2006, by +3.30%. And the DEM registration edge is +1.10% stronger than it was in 2008.

Since its inclusion in the Electoral College in 1964, DC has gone for the Democratic candidate for president every time, and with a blowout margin no less than a whalloping +56.54% (1972, for McGovern in the Nixon landslide re-election). Barack Obama set a percent-and-margin record for DC in 2008: Obama +85.92%, the largest margin for any candidate since 1944 (Roosevelt, MS, Roosevelt +87.12%). DC does not have a congressional delegation, but its mayor is a Democrat and the DC council is overwhelmingly Democratic (11 D – 2 R). DC has the highest partisan registration for one party in any „state“ in the nation. Put bluntly, there are not enough Republicans in DC to even make a dent in the vote. No Republican candidate for President since 1964 has campaigned in DC, if ever.

There is one important difference to WY: while WY had a negative growth rate over 2007 (-20.14%) and over the GE in 2008 (-13.60%), DC has had a equally POSITIVE growth rate over both 2007 and the GE 2008 (+16.73% and +7.68%, respectively). And DC has more than double the number of registered voters as WY. It also has more registered voters than the number of total votes in WY in any one election in its history. At the end of this report you will see yet another surprise about DC.

In the cases of both WY and DC, the voter registration, registration trends and voter history all indicate that they are 100% locks for their respective party.

Category II: stong partisan edge, strong unaffiliated percentage

RI is the state with the next highest DEM registration edge - a blowout +30.72%, but in this case, the unaffiliated voters make up the plurality of the state (48.74%). Assuming a very likely error in the data from LA from 2007, then RI had the largest growth rate of any state by a four year comparison. This state has shown remarkable stability through many years: it has hung around the +30% VR edge for Democrats for a long time. Technically, on paper, RI trended GOP by +0.73% (represented as DEM -0.73% in the graphic) over 2007 and by +0.88% since the GE of 2008, but a closer look shows that both parties lost in percentage over both 2007 and 2008 and the unaffilliated ranks gained almost 2% over 4 years ago.

Rhode Island's electoral history tells a similar story: it has gone for the Democratic candidate for President for 11 of the last 15 cycles, back to 1952, and in 17 of the last 21 cycles, all the way back to 1928. Both Nixon and Reagan captured RI in their re-elections, but with single digit margins, ditto Eisenhower in his first election in 1952. Eisenhower's 1956 re-election was the only time in the last 84 years that a Republican won this state with a double digit margin (Eisenhower +16.52%, 1956). The congressional delegation from RI is completely Democratic. The Governor is an Independent (Chafee, a former liberal republican), the SOS is a Democrat, the Rhode Island Senate (29 D – R 8) and Rhode Island House of Representatives (65 D - R 10) are both overwhelmingly democratic. RI and MA have long traded 1st and 2nd places for the most Democratic state in the Union (excluding DC), ala WY, OK and UT for the Republicans.

RI is a good example of a state with an unaffiliated plurality but such a large partisan edge, in this case a Democratic registration edge, plus a demonstrable tendency on the part of the unaffiliateds to vote Democratic in the election, so that this state is very much a „lock“ for the Democratic Party. With only 10.24% GOP registration here, the Republicans must practically sweep the entire table with the unaffiliated vote in order to win statewide. Specifically, they need more than 82% of the unaffiliated vote in order to come over 50%, assuming the normal small voter swaps between parties at election time.

AZ is an example of a state where the two parties and the unaffiliated voters are relatively evenly split. Here, the edge is GOP +4.20%. AZ also had a major growth rate over 2007 (+23.01%) and also a strong growth rate over 2008 (+7.44%). As in the case of RI, technically, this state is trending for the other party (the Democratic Party, by +1.94%, represented as -1.94% in red), but just as in RI, both parties have slightly lost on registration percent and the unaffiliateds have gained. Over 2007, the GOP has lost more in percent. Over 2008, the DEMS have lost more in percent. But either way, it has remained a modest GOP edge in this state in which both parties must rely heavily on the unaffiliated voters to win elections.

AZ's electoral history is pretty much a mirror image of RI's or MA's, only stronger: from 1952 to 2008, it went for the Republican candidate 14 of 15 times. The congressional delegation from AZ is majority GOP: 2 GOP senators, 5 of 8 representatives are GOP. The Governor and SOS are Republican. The Arizona House of Representatives (40 R – 20 D) and the Arizona Senate (21 R – 9 D) are both overwhelmingly Republican controlled. True, hispanic immigration and the immigration debate that came out of SB-1070 in AZ are changing the demographics of this state and President Obama can certainly try to make a play for this state and perhaps turn it into a major battleground for 2012, but the VR and the electoral history clearly stand on the side of the Republican Party in AZ.

I would be remiss were I to not specifically mention the largest state in the Union, CA. There are some similarities between CA and AZ: both have an advantage for one party and a substantial amount of unaffiliated voters, but the mix is different here: in CA, the DEM edge is currently +13.16%, a major (+8.32%) shift over 2007. And at
44. 04% of the RV, CA democrats are not far from majority status. Assuming they pick up a couple of GOP and third party voters in an election, they only need less than half of the unaffiliateds to win easily, whereas in AZ the GOP needs the vast majority of unaffiliateds to win. One thing that stands out about CA that I will be pointing out about a number of states is its stability compared to 2008: the stats have only slightly moved since then; the DEMS have gained +0.13% on registration edge over the GOP since Barack Obama was elected and the state has experienced negative growth rate of only -0.68% following a historic General Election like 2008. Both of these changes are statistically insignificant. Facit: the Democratic Party still holds a very strong hand in VR in CA, which was proven in the 2010 mid-term elections.

CA's electoral history is much more mixed: since 1932, it has been an even 10-10 for each party out of 20 cycles and it has been more „battlegroundy“ than people think: in 1960, CA was first called for Kennedy and then the call was corrected the next morning (remember Florida 2000? It wasn't the first time...); Nixon won his home state by only +0.55% in 1960 and by only +3.08% in 1968. Ford barely held this state in 1976, with +1.78%. But the trend to the Democratic party could already be seen in 1984, where Ronald Reagan won his home state with a slightly less margin than his first election, which means that his own state trended away from the national pull in the middle of a major landslide election. And it was a close race between Bush 41 and Dukakis in the Golden State in 1998 (Bush +3.57%). Since 1992, however, CA has gone reliably Democratic in presidential cycles, and with no less than a +9.95% winning margin (Kerry, 2004). Obama's +24.03% margin here is 2008 is the largest for any candidate in CA since 1936. California's congressional delegation is overwhelmingly democratic: 2 senators and 34 of 53 representatives. Both chambers of the California assembly are firmly in democratic hands: the California State Senate (25 D - 15 R) and the California State Legislature (their term for the State House of Representatives, 52 D - 28 R). The last Republican Governor of this state, Arnold Schwarzenegger, now has a -55% approval rating (20% approve / 75% disapprove).

CA is also a majority-minority state, like NM, HI and as TX has now become.

Quite obviously, massive changes in VR in this state could lead to a re-evaluation of its expected performance in the next GE, but if conditions (VR, polling, demographics) remain as they are now, then CA will not be competitive in 2012.

Category III: surprising shift in VR

First, when you go to the Ohio SOS website, it will be very hard for you to find any VR statistics. Ohio is the ONLY state in the Union to actually provide .txt lists of all registered voters in the state, by party affiliation, county and address. You can download the list (as .txt and import as .csv) per county or statewide, but there are so many fields that a normal EXCEL-like program cannot handle the number of delimited fields. The large list ist 200 MB large!. And the lists do not contain grand totals, so I contacted the Ohio SOS office and they sent me the data for the time frames I requested. The results were shocking. Ohio has gone from an unaffiliated majority state in 2008 (54.20%) and 2007 (64.86%) to an unaffiliated blowout state in 2011: 78.51%. Both the Democratic and Republican parties have taken a hit here, but the Democrats have taken the heavier hit. What was a DEM +10.57% registration edge in 2008 and a +2.21% registration edge in 2007 is now a GOP +0.83% edge, but both parties are so in the low teens that anything is possible. This is a jump of +13.64% for the unaffiliateds over 2007 and a +24.31% shift to the affiliateds over 2008. This means that since the GE 2008, 1/4th of the Ohio electorate has turned away from the major parties. The third party vote cannot be properly tallied as it was not in the stats in the same way before 2010, but even if 2% of that 24.31% shift is true third-party vote, then that is still a massive shift.

I am sure that both the Obama and the GOP statisticians have this data. Anyone who wishes to have the data from me can receive a forward of the email I received from the Ohio SOS. Just send me an email. This makes Ohio extremely volatile looking forward to the 2012 election. Perhaps for this reason the Obama team has already announced (as I had predicted on my blog on a number of occasions) his intention to make a strong play for GA, SC and AZ on top of MO.

Why? GA has 16 EV. OH has 18 EV. In 1992, GA had 13 EV, OH had 21. In 1976, GA had 12 EV, OH had 25. In other words, in 1976, Ohio had double the electoral firepower of Georgia. Now they are pretty much even and it would not surprise me if GA moves up to 17 electors while OH moves down to 17 electors in 2020. So, if Obama, who lost GA by just -5.20% in 2008, picks-up the state in 2012, but loses Ohio, that is a net loss of only 2 electors.

Concerning the electoral history of my home state, I already did an exhaustive 3-part analysis, which you can read here: OHIO - Part I, Part II, Part III , raw data

Having now seen excerpts from the data, here is once again the map from above, with the partisan advantage in 2011 and the shift over 2007 marked in for each state (this is the data from the EXCEL document):

Here the data in a more logical format from the excel worksheet.

Here the states, in descending order of partisan advantage:

19 of 30 states have a +10% or more partisan advantage, from DC down to AK. The higher the advantage, the more likely the „lock“, but not in all cases (LA, KY, OK, WV)
Not surprisingly, a number of states with narrow partisan advantages are also the states that are often battleground states in elections.

Here the same list, sorted by the unaffiliated voters, in descending order:

Ohio is at the top of this list. We see that it is clearly possible to have an unaffiliated majority state and still think of that state are very red or very blue (see: AK an MA right under OH), but 78.51% unafilliated puts Ohio way in front of all the rest. Also, the bottom of the list is interesting and a problem for the GOP: PA has the smallest percentage of registered unafilliated voters (6.22%) and the Democratic party already has a +13.78% edge.

Once more, this list, organized by GOP %, descending:

We see that where the GOP has a majority, it has a solid lock.

Here that list by DEM %, descending:

We see a larger number of states with DEM majority VR registration. But that, as I wrote at the beginning, is not predictive: look at WV, OK, KY.

One last time, the list organized by the states' percent share of all registered voters from 49 out of 51 „states“:

Here we can see where the biggest prizes are. The list sorted in this fashion is very, very helpful. Of the top ten states in terms of size within the electorate at this time, the top ten states, from CA to CO comprise 44.08% of all registered voters in the USA at this time. 6 of those 10 states were Obama landslides in 2008. One was a near landslide (CO), two were lean wins (OH, FL) and one was a squeaker (NC). First at the 11th state (AZ) do we have a GOP win from 2008, which then brings the list to 45.89% of all RV in the USA at current. Remember, these 30 states together make up 59.32% of the registered electorate right now, so the other 19 states on this list combined represent 13.43%. Also, please notice that DC not only has more registered voters than WY, it also has more registered voters than IA, which has 6 Electoral Votes, while DC has only 3.

The only thing I have not analysed yet are the smaller shifts in partisan advantage for the states where this is happening, for I wish to have at least 2 more quarters of research to back it up. There are some other details I will be analysing, but first I thought to let me readers sort some of this out with their own eyes.


Credits to the following individuals who helped: Gregory McBurney and Rob Rock from the Rhode Island elections division, and Robin Fields, Election Assistant to the Ohio Secretary of State.

1 comment:

  1. http://www.sos.state.or.us/elections/pages/history/stats/11mvr.html

    Oregon 2011 can be found here.


Constructive comments and critique are always welcome. Please keep it polite and respectful.