Saturday, December 31, 2022
Scratch The Surface and You Discover Something Else: The Year of Biden and Gas Prices in 2022
A project that started in curiosity, for the past 1,776 days and weeks, I have recorded data points on daily gas prices at the Big Smoke gas station at 233 W. Franklin Road in Meridian, and the daily approval rating of U.S. Presidents Trump and Biden from either RealClearPolitics or FiveThirtyEight polling aggregators.
In 2022, we have an instance of a seemingly concrete beginning and ending masking a lot of negative drama in the middle. On New Year's Day, President Biden's approval rating from FiveThirtyEight's polling aggregator was 43.3. On New Year's Eve, today, his approval is 43.3. Of course, this is a misleading conclusion about a totally static approval rating, the depths of which were reached in latter July, 37.5 on July 21st, a little after gas prices reached their zenith, $5.29 on June 29th. This suggests a bit of causal action, with gas prices a leading indicator of Biden's approval travails.
What turned things around? Gas prices backed off quite a bit from their high, and rather quickly, 34 cents lower on July 29th than on June 29th. And Biden's approval inched higher once a key event occurred: Senator Joe Manchin and Senator Chuck Schumer struck a deal on the Inflation Reduction Act, marking its eventual signing into law on August 17th. By then, Biden's approval had inched upwards to 40.4, with that three point jump being a rather strong move for a relatively static indicator all year long (standard deviation of 1.29). The legislative dam for this signature, $770 billion, bill had been broken, and Congress could legislate spending items that Democrats argued would affect all Americans positively, while Republicans argued the opposite.
This latter half of the year was an electoral bet by both parties that either inflation and big spending would get the thumbs-down from the electorate (the Republican bet, of course), or a productive Congress (one of the most productive in generations, on razor-thin majorities in the House and Senate) would be seen favorably (the Democratic bet). In the end, inflation backed off enough from its summer high to reduce the potency of the spending issued for the Republicans, and combined with two non-economic features of the political landscape (the overturning of Roe v. Wade and novice Republican candidates for key Senate races), the Democrats made some historic gains in the midterm elections (all Democratic incumbents winning their re-election races, and the president's party gaining gubernatorial mansions for the first time in a midterm election since 1934).
This latter half does overshadow the alarm of the first half of the year, when inflation spiked to 9.1 percent in June and the war in Ukraine added a jolt to global energy markets, all contributing to massive prices at the pump: from $3.49 on February 20th just prior to the Russian invasion on February 24th to $5.15 on June 24th, a 48% increase. There was little cause for optimism in these figures for President Biden. But the Pearson Correlation of Biden approval and gas prices for the entire year was pretty consistent throughout, and ended at -.69, a very strongly negative relationship. If prices backed off at the pump, Biden's approval rose, and vice versa.
Will this pattern continue in 2023? It is very hard to say. On the one hand, the Pearson Correlation for Trump's final two years in office was very slight: .14 in 2019 and -.18 in 2020. Yet the correlation has been very strong for Biden the past two years: -.69 in 2021 and -.69 in 2022. Given 707 data points for Biden, I will bet on big data. Biden and the gas pump appear to go together.
Thursday, December 23, 2021
A Fractured Electorate? French Presidential Election Forecasting for 2022
Ross E. Burkhart
School of Public Service
Boise State University
23 December 2021
France will experience its every-five-years presidential election next April. Political change has been the constant across France and its neighbors over the past five years, from the gilets jaunes unrest and COVID-19 vaccine protests in France itself to completion of the British Brexit to the conclusion of Angela Merkel's tenure as German Chancellor. Add to this general upheaval across the European Union regarding migration, economic decline, populist rise and the spread of COVID-19, and the chaotic continental political pattern continues.
Focusing on France, the destruction of the Fifth Republic establishment parties during the 2017 presidential election remains so. The Parti Socialiste candidate for the presidency, Paris Mayor Anne Hidalgo, received 5% support in the November public opinion polls and the Républicains candidate, Valérie Pécresse, received 14% support, well below the party standards from past election campaigns. Even the standard-bearer of the increasingly ascendant Rassemblement National (formerly National Front), Marine Le Pen, slipped below 20% support, with the slack taken up by the rank outsider on the extreme right, Éric Zemmour and his Reconquête movement.
Strangely enough, the most constant aspect of the current presidential campaign is the enduring 35-40% support for the insurgent 2017 presidential winner, Emmanuel Macron. An avowed centrist (and former Socialiste), Macron has studiously attempted to straddle the middle in navigating his presidency's way through a thicket of issues, from existential (the response to COVID-19) to identity-based (secularism). These political efforts have gained him at a modicum a grudging respect for surviving during these treacherous times, though there is the suspicion among many that rather than being a principled centrist, he instead is diving toward the center for the true Downsian median voter that in theory could propel him to a triumphal re-election.
How is Macron doing toward his quest? So far, reasonably well. The November polling showed him about ten points ahead of his nearest competitor, Le Pen, 28% to 18%. Of course, this level of support is nowhere near the majority support required to end the election at one round. Macron's strategy is surely to hew to the median center, survive to the second ballot, and receive the Arrowian voter second preferences on the second round to return to the Elysee Palace.
Of course, doing well today is no guarantee of success tomorrow. Hence, my French election forecasting effort attempting to foresee the election results. Is it possible to forecast French elections? The evidence suggests that it is possible. Stretching back to the pioneering work of Michael Lewis-Beck (1991), despite the small number of Fifth Republic presidential elections, they are remarkably well-behaved from the standpoint of hazarding a forecast of the impending election.
Visual evidence of this is from the figure below. (This even includes the extreme outlier election of 2017, during which the French political world was irrevocably rocked.) Along the horizontal axis are data of approval of the French president, six months prior to the election, stretching back to the 1965 election. Along the horizontal axis are data of the first-round vote for the presidential candidates on the ideological Left. (To make this graph work properly, the approval data are adjusted in the following way: for a president of the Left, the data are as is, while for a president of the Right, the data are 100 minus the approval data.) There is a definite upward slope to the data points, indicating a strong relationship between presidential approval and vote. This is confirmed by the Pearson correlation between incumbent popularity and Left vote, r = .84, N = 10.
To forecast an election result, three things are essential: a good theory of voting, an accurate forecast, and significant lead time for the forecast. A reasonable voting theory to consider for forecasting is the "reward and punishment" theory first put forth by the Harvard political scientist V.O. Key in his book The Responsible Electorate (1966). In this scenario, voters are the "vengeful gods" who will vote for a government that performs well and against a government that does not. It is a relatively broad calculation for the voters to make, but the evidence suggests that make it, they do (Kramer 1973, Tufte 1975, Lewis-Beck 1988, Stimson 1991).
If presidential approval is a good proxy variable for government performance, and the literature suggests that it is, then presidential approval with appropriate lead time prior to the election should serve as a strong cue to the voters on Election Day. The approval data, six months prior to the election, for 1965 – 2007 come from Nadeau, Bélanger, and Lewis-Beck (2012), and I have collected the data for 2012 and 2017.
A basic model proposed by Nadeau, Bélanger, and Lewis-Beck (2012) suggests itself, to be estimated with ordinary least squares regression:
Left Vote = a + b Incumbent Approval
The estimates are as follows:
Left Vote = 24.25 + .38 Incumbent Approval
(5.97) (4.80)
R-squared = .77 S.E.E. = 4.10 Durbin-Watson = 2.01 N = 9
Where the variables are measured as stated above, the figures in parentheses are absolute t-ratios, the R-squared is the percentage of variance explained in the Left Vote DV by the Incumbent Approval IV, S.E.E. is the error in prediction of the DV by the IV, Durbin-Watson is the Durbin-Watson statistic for autocorrelation, N is the number of elections in the dataset, and the estimation technique is Cochrane-Orcutt AR(1) regression via Stata.
For a forecasting model, the R-squared value is of primary importance, as the model must account for the vote as precisely as possible. R-squared values above .70 are generally acceptable models for an accurate forecasts, with the higher the value, the more accurate the forecast model.
With this model's R-squared of .77, this model is sufficiently strong enough to render an accurate forecast.
What is the forecast? We simply plug in the relevant current value for Incumbent Approval and solve the equation. The relevant approval for Emmanuel Macron in November 2021 is 43 percent. This approval is on the lower end of presidential approval ratings, but it is certainly not catastrophic as in was in 2017.
The toughest call here is to characterize Macron: is he of the political center or of the left? His conduct while residing in the Élysée Palace suggests a centrist hue to his tenure in office. He was the economics minister in the Hollande government, so he also has some left credentials. On balance, his ideology is on the center-left, not the right, so we can declare his incumbent popularity for the left vote.
To solve the equation, then, the results are as follows:
Left Vote = 24.25 + .38 (43)
Left Vote = 24.25 + 16.34
Left Vote = 40.59% of the first ballot vote
This tracks pretty well with the current levels of public support for the left candidates plus Macron, who together add up to around 45 to 49 percent of support. Despite all of the clamor associated with the far right, the forecasted vote for candidates on the left during the first round of balloting is likely to be substantial.
References
Lewis-Beck, Michael S. 1991. "French National Elections: Political Economic Forecasts." European Journal of Political Economy 7:487-96.
Lewis-Beck, Michael S., and Tom W. Rice. 1992. Forecasting Elections. Washington, D.C.: CQ Press.
Nadeau, Richard, Éric Bélanger, and Michael S. Lewis-Beck. 2012. “Proxy Models for Election Forecasting: The 2012 French Test,” French Politics 10:1-10.
Stimson, James A. 1991. Public Opinion in America: Moods, Cycles, and Swings. Boulder, CO: Westview Press.
Saturday, September 05, 2020
Family on Father's Day (sans wife who is the photographer)
Tuesday, December 31, 2019
That Was Quite the Mess
I'll end the year with a look back at a failed model, namely the forecasting model for the recent British elections. It was a pretty spectacular failure at 75 seats short of the actual result. But what did Edison say, supposedly...it's not failure, but just another example of what doesn't work. In the spirit of "we write to learn" (many folks have said that), I present the following essay.
What Went Wrong? A Post-Mortem of an Election
Forecast
Ross E. Burkhart
School of Public Service
Boise State University
14 December 2019
Yesterday,
the aftermath of a consequential election in a critical Western democracy, the
United Kingdom, came upon an important day for our Political Science students:
the capstone course project showcase. A brief recap of both events follows, and
then the theme of this short essay emerges: reflecting on a poor election
forecast.
The
election process is simple in Britain: voters tick a box for their favored
candidate and place the paper in a box, to be counted by the returning
officer's staff in their geographic constituency. The winners in the 650
constituencies, decided by plurality (meaning winning at least one more vote
than their closest opponent), are thus duly elected Members of Parliament. The
party winning the most seats gets to form the next government. In this case,
the Conservatives won a smashing majority of 80 in the House of Commons. This
is the Conservative Party's largest majority since 1987, and suggests a
fundamental realignment in British politics, where working-class voters crossed
over to the Conservatives who promised a speedy Brexit to address working-class
grievances over being ignored by globalized Brussels bureaucrats.
Meanwhile,
the capstone Political Science students presented their projects in poster
format, on public display in the Student Union Building. They conducted
research on an impressive array of topics: Title IX funding and transgender
rights in schools, policing reform and civil liberties, increasing the
intelligence budget, Twitter followers and campaign spending, balancing
abortion politics, how reporters cultivate relationships with politicians, the
prospects for a hard or a soft Brexit, the effects of door-to-door campaigning
in state legislative elections, water pollution in Costa Rica, state-level
abortion restrictions and the Supreme Court, and human rights violations in
Burma and China. The students also had an eye in their poster presentations
toward highlighting findings and shortcomings in their research. As critics,
they were most effective evaluators in pointing out the shortcomings in their
research and suggestions for future refinement. Such impressive self-reflection
behooves me to follow their example with one of my recent research pieces.
Two
weeks ago I posted "A Snap Elections Forecasting Model of British
Elections" in which, using data that reflected the snap election
environment (defined as an election occurring well before the scheduled end of
the Parliament), I gave a forecast of the Conservatives winning 290 seats in
the House of Commons. It was a mere 75 seats off! I need to understand what went wrong with the model. The
forecasting model was as follows:
S =
20.50 + .73 P + 1.61 J – 3.78 T (Eq.1)
(1.45) (3.03) (2.74) (1.82)
R2
(adj.) = .68 S.E.E. =
5.42 D-W = 1.84 N = 15
Where
S = percentage of House of Commons seats won by the incumbent governing party,
P = approval of the performance of the Government at either the time of the
dissolution of Parliament during a snap election or six months prior to
dissolution during a non-snap election, J = jobless rate at the time of
dissolution during a snap election or six months prior to dissolution during a
non-snap election, T = the number of terms the incumbent party has been in
office assuming a fatigue factor takes place. The figures in parentheses are absolute t-ratios, S.E.E. is the standard error of estimate, and D-W is the Durbin-Watson statistic after the Cochrane-Orcutt procedure.
I
noted in my forecast that the R-squared statistic is on the low side for a
forecasting model, and that the standard error of estimate is rather high,
suggesting a fair bit of unexplained variance in the seats forecast, though
certainly nothing like the 75 seat error in forecast. Of course, that much
error is unacceptable in a forecasting model, hence this postmortem.
We
are left with several possible paths toward enlightenment. One is that the
election was an outlier, a freak of nature that could not be forecasted well.
We reject this path because other forecasters were rather close to the mark
(Murr, Stegmaier, and Lewis-Beck 2019), but there were prominent forecasts that
were wide of the mark (Lebo and Fisher 2019), suggesting there was no real
unanimous view as to the forecasted result (Greenwood 2019).
A
second possibility is that British elections are inherently unpredictable, with
those who claim to make a forecast very close to the final result just having
been lucky in their forecasting. After all, forecasting is an informed
probability exercise, and even a broken clock is correct twice a day, goes the
old adage, so anyone has a chance of getting it right. While there are critics
of the election forecasting enterprise (van der Eijk 2005), this possibility of
unpredictable elections seems unlikely as well. Many elections are well
forecasted (Burkhart 2018). More broadly speaking, the forecasting literature
has successfully forecasted election results in many countries over a lengthy
time period, too well to ignore (Bélanger and Trotter 2017).
A
third possibilithy is that the model itself is rather faulty. There are three
pieces of evidence that point in this direction. First, the R-squared is rather low for forecasting purposes
at .68. Ideally, we seek an R-squared of at least .75 in election forecasting
models. Second, the unemployment independent variable is oddly signed in the
positive direction, suggesting that incumbent governments that raise
unemployment rates are more likely to win more seats. (The positive sign is
most likely influenced by Margaret Thatcher's three strong majorities during
the highest unemployment rates in British history since the Great
Depression.) Third, the
statistical significance of the terms in office variable is pretty marginal at
approximately the .09 level. The preferred significance level is .05.
Further
examination reveals that the next election forecast, presumably taking place in
2024 given the substantial Conservative majority, erodes in model performance.
The forecast equation for the 2024 election is as follows (Cochrane-Orcutt
estimation):
S =
7.16 + .95 P + 1.68 J – 1.37 T (Eq. 2)
(.56) (4.10) (3.45) (.78)
R2
(adj.) = .61 S.E.E. =
5.60 D-W = 2.06 N = 16
All
model performance indicators decline in 2024 compared to 2019. The y-intercept
is no longer significant, indicating that the baseline percentage of seats for
the Conservatives, while estimated to be 7.16% of all seats, could actually be
0%. The popularity function is elevated in importance, almost equating a
one-to-one impact on the Conservative vote. The jobless independent variable
remains steady in impact, while the trend variable drifts into statistical
insignificance.
In
fact, simply dropping the trend variable yields the following result:
S =
-.63 + 1.07 P + 2.35 J
(Eq. 3)
(.27) (5.53) (2.95)
R2
(adj.) = .62 S.E.E. =
5.51 D-W = 2.15 N = 16
This
model performs better than Eq. 2 in that all variables are statistically
significant and the adjusted R-squared rises, if just a smidgen. Yet the
y-intercept suggests that the baseline for Conservative seats is effectively
zero, which is nonsensical.
To
be sure, more work is needed in understanding a proper structural model of
British election forecasting. In the face of the strong performance of the
citizen forecasting models and the multilevel YouGov model, it is tempting to
proclaim this the elegy of the structural equation model. However, we must
remember another statistical issue with this literature, and that is small
sample sizes. The Central Limit Theorem is a good rule of thumb for regression
analyses upon which these forecasts rely. The 2019 election is one data point
out of the small sample of 17 British general elections in this model, so
perhaps the only elegy worth making is that it is too soon to tell its demise.
REFERENCES
Bélanger, Éric, and David Trotter. 2017. "Econometric Approaches to
Forecasting." In
The
SAGE Handbook of Electoral Behavior,
eds. Kai Arzheimer, Jocelyn Evans,
and
Michael S. Lewis-Beck. Beverly Hills, Calif. Sage Publications.
Burkhart,
Ross E. 2018. "Was Anyone Right? Assessing the 2016 U.S. Presidential
Election
Forecast." The Blue Review,
15 January.
Greenwood, Joe. 2019. "Experts predict that the general election will be tighter than
Greenwood, Joe. 2019. "Experts predict that the general election will be tighter than
expected."
LSE British Politics and Policy,
10 December.
Lebo,
Matthew, and Stephen Fisher. 2019. "Forecasting the 2019 General Election
using
the
PM and Pendulum Model". LSE British Politics and Policy, 20 November.
Murr,
Andreas, Mary Stegmaier, and Michael Lewis-Beck. 2019. "Citizen
forecasting
2019:
a big win for the Conservatives." LSE British Politics and Policy, 4
December.
van
der Eijk, Cees. 2005. "Election Forecasting: A Sceptical View." British
Journal of
Politics
and International Relations 7:210-4.
Friday, November 29, 2019
A Snap Election Forecasting Model of British Elections
Here is a brief research exercise I undertook out of interest in both the concept of snap elections in the Westminster parliamentary system and election forecasting in general. It is not of sufficient length and depth to publish in a conventional space, hence I am making it available here. Enjoy!
*****
The Snap Election Forecasting Model of British Elections
Ross E. Burkhart
School of Public Service
Boise State University
I make an election forecast of the 12th December 2019 British general election based on a common structural model of elections, with a small innovation that accounts for snap elections. The forecast suggests that the Conservative government will lose seats, but with a larger number of parties than normal likely to win seats, the governing party or coalition that will emerge from the election is uncertain.
This particular British election may very well be hazardous to forecast, given the controversial circumstances surrounding its calling, the high number of parties which will be contesting the constituencies with plausible chances of victory, and uncertainty over how the voters will weigh the primary issue, Brexit. (To call this election the "Brexit Election" is almost an understatement!) Structural forecasting models that rely on "heavy" traditional variables to make a forecast, such as economic and governmental performance, may well go awry given the special nature of this election. Yet forecasters cannot be choosy and remain credible. The election is upon the country in less than a month's time.
For context, the structural models are pioneering British election forecasting forays. They have been joined recently by aggregate polling models from several polling firms and synthetic models combining the two approaches. The structural models employ economic and political variables measured with a good lead time before the election date, in order to generate a forecast well beforehand. Forecasting models are evaluated under three criteria: (1) sound theory behind the forecast, (2) non-trivial lead time for the forecast, and (3) model performance. Structural models that invoke common theories of voting (referendum voting à la V.O. Key, voter fatigue à la Alan Abramowitz) suggest an external validity to the forecasting enterprise. Combined with sufficient lead time, and good model performance, structural models play an integral forecasting role.
All of these forecasting techniques presuppose a similar cognitive voter dynamic. The election dates are usually known months in advance, allowing voters the time to ready themselves for evaluating the parties and candidates. But what about snap elections? These occur in the Westminster parliamentary system for three reasons: (1) the government loses votes of confidence in the House of Commons, (2) the Prime Minister seeks to take advantage of her or his popularity to win more House seats, (3) the government is stymied prior to the end of Parliament in passing legislation due to deadlock in the House chamber, often caused by an attrition of seats held due to defections, resignations, and death.
Snap elections characterize the current election campaign, and several other prior national-level general elections over the past 60 years. These elections are listed below, along with the date of dissolution of Parliament and the election date itself. (The dissolution of Parliament rapidly follows the declaration of the general election by the Prime Minister.)
TABLE 1: BRITISH SNAP ELECTIONS
Year
|
Dissolution of Parliament
|
Election Date
|
1955
|
7 April
|
26 May
|
1966
|
10 March
|
31 March
|
1974 (Feb.)
|
8 February
|
28 February
|
1974 (Oct.)
|
20 September
|
10 October
|
1979
|
7 April
|
3 May
|
2017
|
3 May
|
8 June
|
2019
|
6 November
|
12 December
|
From a practical standpoint, it makes sense that voters take stock of the condition of the country at the time of dissolution for a snap election, as while there may have been buildup to the dissolution, by definition there is some surprise to the declaration timing itself. Hence, voters should feel uncertainty in their evaluation, and uncertainty breeds a quick and dirty evaluation, right at the time of the election. Hence, the election forecast will have little lead time in the snap election environment, but conditions six months prior seem out-of-date.
I re-create a structural British elections model forecast equation from the extant literature, with the innovation of utilizing data from the dissolution date for the snap elections. The modern structural modeling involves elections and data beginning in 1955, which is where I begin my dataset. The model is informed by Lewis-Beck, Nadeau, and Bélanger (2004), where the variables are as follows: percentage seats won by the incumbent party as the dependent variable (denoted as S below), and three independent variables: the public approval rate of the Prime Minister or P (either six months prior during non-snap elections or at the time of dissolution during snap elections), the jobless rate six months prior to the election or J, and the number of terms in office for the incumbent party assuming a fatigue effect for long-running spells in power for one party or T. The functional form of the model is thus: S = f (P, J, T)
I estimate the model using a Prais-Winsten regression, Cochrane-Orcutt procedure (Stata statistical software) in order to account for autocorrelation. The model results are as follows:
S = 20.50 + .73*P + 1.61*J - 3.78*T
(1.45) (3.03) (2.74) (1.82)
R2 (adj.) = .68 S.E.E. = 5.42 D-W = 1.84 N = 15
Where R2 (adj.) = adjusted R-squared statistic, S.E.E. = standard error of estimate, D-W = Durbin-Watson statistic for autocorrelation after the Cochrane-Orcutt procedure, N = number of elections, figures in parentheses = absolute t-ratios.
Model performance is good, if surprising. All independent variables reach conventional levels of statistical significance. The popularity and terms independent variables are signed in the expected directions: more popular prime ministers win more seats, and too lengthy a stay in government leads to seat declines. Strangely, the unemployment rate is positively connected to seat increase for the incumbent party. This may seem perverse, but perhaps is reflective of a generally high level of unemployment in Britain as the time period sampled went on, and may also suggest electoral forgiveness for high jobless rates if they trade off for lower levels of inflation and general income growth. (A technical point: there is no significant multicollinearity among the independent variables, with the highest Variance Inflation Factors (VIFs) being 1.5.)
The forecast for 2019 is made simply by plugging in the relevant measurements for 2019, with the snap election polling at the time of dissolution revealing 40 percent approval of the Prime Minister, Boris Johnson.
S = 20.50 + (0.73*40) + (1.61*3.8) - (3.78*3)
S = 44.48
.4448 * 650 = 289 seats won by the government party
This forecasted result does not suggest a resounding victory for the Conservative government at the polls, rather a loss of nine seats after dissolution, but there is more than the usual uncertainty surrounding this particular snap election. (Witness the high standard error of estimate and relatively lower R-squared statistic.) While at once the Conservative government has gained in popularity in recent months, its lengthy stay in power (since 2010) may find the fatigue factor pushing voters toward the opposition. Low levels of unemployment actually give this variable the least impact on seat change. However, with several parties poised to win seats across the country, 289 seats may well be enough for the Conservatives to form the next government.
Wednesday, January 03, 2018
End of the Project
With a new year at hand, all good projects come to a close, and this one is a forced ending, in that the Gallup polling organization will no longer issue daily presidential approval ratings. Instead, it will switch to weekly presidential approval ratings. I will continue to gather data on a weekly basis, but that will yield a much smaller sample size and miss the ups and downs of prices and approval. For now, here's the data analysis. Enjoy!
Pearson correlation of daily gas prices and President Trump approval from 1 April 2017 through 30 December 2017 = -.324, N = 343. As President Trump's approval increased, gas prices modestly decreased, though this is most likely coincidental with the seasonal dip in gas prices.
Monthly Pearson correlations:
April = .425
May = .417
June = -.037
July = .102
August = -.535
September = -.146
October = .651
November = .262
December = -.675
This shows the monthly volatility in gas prices. President Trump's approval has been on a gradual decline over the months.
Monthly average Trump approval:
April = 40.633
May = 39.806
June = 37.933
July = 38.065
August = 35.867
September = 37.333
October = 36.667
November = 37.500
December = 36.333
Sunday, August 06, 2017
A Summertime Project
UPDATE: 22 NOV 2017: To my surprise, there is a small literature on local economic conditions such as gas prices, overall economic evaluations, and political outcomes such as presidential voteshare. The literature includes: Abrams & Butkiewicz, 1995; Ansolabehere, Meredith & Snowberg, 2013 & 2014; Books & Prysby, 1999; Brunk & Gough, 1983; Reeves & Gimpel, 2012; Weatherford, 1983.
UPDATE, 28 OCT 2017: The Pearson correlation between Trump approval and the Big Smoke regular gas price over the past 210 days is -.45, a moderate level.
I have been creating a long-term data set of two variables: President Trump's daily approval rating as published by Gallup (http://www.gallup.com/poll/201617/gallup-daily-trump-job-approval.aspx) and the daily price of regular gas at my local gas station, the Tesoro Big Smoke station on Franklin Road between Meridian Road and Linder Road in Meridian. I started this project at the beginning of April and will continue it into the foreseeable future.
I have to confess that this is a correlative project and not particularly causal in nature. I do not see an explicit link between President Trump's actions in Washington, D.C. and the price of a basic commodity in distant Idaho. (My area is called the most geographically isolated urban area in the U.S., a full 5 1/2 hours away in driving time from the nearest urban center, Salt Lake City.) Yet both fluctuate, Trump on a daily basis as rarely does he retain the same percentage support day after day, and gas on a weekly basis as prices bounce up and down based on supply and demand.
So far, the correlation from April 1st until today is slight: the Pearson correlation is = -.19, meaning that as prices for gas increase, the president's popularity decreases, but by a very slight amount. Much of this relationship surely can be accounted for by factors other than any direct linkage between the two. For instance, it is well known that as demand for gas increases during the summer vacation and driving seasons, prices will increase, and this summer is no exception, rising from $2.35 a gallon in April to $2.55 today. The president's approval rating has been in a gradual descent since the spring, with his average daily approval rating in April being 39% while in July it was 38%.
If nothing else, this is a good exercise in restating the dictum that correlation is not the same as causation!
