Models by benthamite

// Assumptions
total_campaign_benefit = beta({mean: 0.07, stdev: 0.04}) / 2 // div by 2 because we are only increasing half the voters' turnout
labor_share = beta({mean: 0.01, stdev: 0.01})
labor_benefit = total_campaign_benefit * labor_share
default_margin = normal(-0.07, 0.07)  // dan osborn's margin was -7%

// Calculations
n = 10000
labor_samples = sampleN(labor_benefit, n)
margin_samples = sampleN(default_margin, n)

// Assumptions
total_campaign_benefit = beta({mean: 0.07, stdev: 0.04}) / 2 // div by 2 because we are only increasing half the voters' turnout
labor_share = beta({mean: 0.01, stdev: 0.01})
labor_benefit = total_campaign_benefit * labor_share
smallest_margin = normal(0.005, 0.003)
election_won_by_one_state = beta({mean: 0.07, stdev: 0.04})

// Calculations
n = 1000
labor_samples = sampleN(labor_benefit, n)

fmap(f) = {
  n = 100
  g(x) = SampleSet.make(List.map(sampleN(x, n), {|z| f(z) }))
  g
}
curry(f, x) = {
  g(y) = f(x, y)
  g
}
fmap2(f, x) = fmap(curry(f, x))

/*
1. Following Gelman et al., I model the vote margin as a normal distribution
1.1. NB: I don't think primary elections are actually well modeled as having normally distributed vote margins? Often there is one very distant frontrunner. But maybe highly contested primaries are well modeled like this and this will be a highly contested primary? caveat emptor
2. Dollars per vote: I loosely based this on a document from Dan Eth about Jay Shooster's campaign but kind of made it up. Better numbers appreciated
3. Primary: 
3.1 Default odds of winning: 40%, per Jay Shooster
3.2 Sd of margin: made up by me saying it will be 10% of total votes
4. General:
4.1 Default odds of winning: Jay suggests that the democrat will win this district iff democrats control the house, and probability for the latter is currently ~75% on prediction markets
4.2 Sd of margin: last year was 2.5k, so I vaguely guessed 5k

/*
Describe your code here
*/

a = normal(2, 5)

/*
Describe your code here
*/

// SAMSHA via perplexity
cocaine_users = {
    'Year': [2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022],
    'Users_Millions': [4.55, 4.8, 5.05, 5.9, 5.5, 5.45, 5.2, 4.8, 5.274]
}

/*
Describe your code here
*/

a = normal(2, 5)

/*
Describe your code here
*/

a = normal(2, 5)

/*
1. Following Gelman et al., model the vote margin in each swing state as a normal distribution, with the mean coming from current prediction markets
1.1. NB: Squiggle doesn't have t distributions, just doing normal for now
1.2. NB2: Prediction market numbers seem sus, I added my own estimate whcih seems more reasonable to me
2. Consider the value of getting N additional voters in a given swing state with total voting population P as increasing the mean of this distribution by N/P
3. Assume that the number of additional voters we generate is 32% of our number of users (per Coppock), and that our number of users is linearly proportional to our budget
4. Sample from this distribution a bunch of times with different budget sizes, and look at the probability that Biden gets >= 270 electoral votes with each budget size
*/

/*
  Estimates advantage of Biden by:

  1. Estimating the probability that proposals of given values will be produced
  2. Estimating the increased likelihood of the proposal being implemented under Biden
  3. Multiplying the above two numbers together

*/

/*
Describe your code here
*/

a = normal(2, 5)

/*
Goal: estimate how many dollars in donations to LTFF are equivalent to one 
hour spent by a CEA employee

LTFF kindly published their marginal grants, and one of them was a (hypothetical)
grant for people to travel to a biosecurity workshop. Fortuitously, CEA staff have
recently spent some time helping with a biosecurity workshop.

This means we can estimate how many dollars the LTFF has to spend in order to 
improve a biosecurity workshop as much as CEA staff improves it by working one hour. 

/*
  ==== Costs of replacing an employee ====

  This model attempts to estimate the cost of replacing an employee.
  Numbers are in months of lost productivity, e.g. if total_cost = 5
  that means that replacing an employee cost the equivalent of 5 months of their labor
  disappearing.

  These numbers are lower than the ones I usually see (6-9 months, https://lrshrm.shrm.org/blog/2017/10/essential-elements-employee-retention).
  I think it's because that study is low-quality and made up  but feel free

/*
Describe your code here
*/

a = normal(2, 5)