Week 5: Making Risk (19 & 20 C)

Question 1

What were the views on management by numbers?

Answer 1

• an integral part of modernity
• undeniably a rising hallmark of the modern state

But many vocal critics

• Karl Marx concerned by the economy’s affect on life
• Max Weber wary of move towards rationalization as bureaucratic structures used modern data
• Jurgen Habermas shared concerns with quantification of life

Question 2

How did data penetrate into public lives in the 19th C?

Why?

Answer 2

Compare pre-19th to post 19th century: it went from a referential point to the core of decision making
o Previously, would rely on an expert to make an informed decision using the data
o Increasingly though, it became purely mechanical calculation based on the input data

-made possible by availability of data, maturity of probability theory, and political culture shift towards “trust in numbers” instead of “expertise”
o E.g. Planning roads with population data rather than just asking a city planner to submit their opinion
• Used algorithmic processes, as they were more objective
o Went hand in hand with the rise of democracy

Macroscopic level: Governments had much more agency in the management of lives (biopolitics) created institutions like Ministry of Health

Microscopic: “The technology of the self”

Question 3

How did US Life Insurance shift in the 19th C?

Answer 3

• originally a philanthropic operation of fraternity
• became a significant business in 17, 18th C as it grew more complicated
• early on could rely on just Life Tables, but needed dedicated actuaries as time went on

US was industrializing in 19th C, many immigrants

• of all profiles, looking for industrial jobs
  
  • so wanted INDUSTRIAL LIFE INSURANCE
  • insurance landscape shifted as old players were upseated
• Civil war ends 1865, slaves liberated

Shifted from trusting expertise to numbers

• Life tables were referential for managers before
• now premiums and payments were purely formulaic

Question 4

How did the Insurance companies deal with the change?

Answer 4

1. Classing
   - clients no longer all middle age white men
   - need to introduce many more factors than age (e.g. geographic region, occupation, medical history)
   - sent physicians to study, and lawyers for credit history to place clients in statistical profiles
2. Predicting
   - use classing + past data on the profile to generate packages
3. Smoothing
   - apply statistical techniques (e.g. curve fitting, interpolation) to remove the real world’s fluctuations from data
4. Prescribing and Preventing
   - most far reaching part of the new insurance
   - death is predictable and preventable
   - emergence of the STATISTICAL INDIVIDUAL
   • a very real person for a data set

Question 5

What was the issue with predicting liberated African American slaves?

Answer 5

Only began insuring them in the postbellum years (i.e. after the civil war)
• But how to set their premiums?
o All data was from antebellum records
o Slavery painted a very grim picture with extremely high mortality rates
• Insurance company’s compromise: Premiums can’t be set higher (as they don’t have the wages to pay significantly more), but benefits offered were a fraction of what was given to other ethnicities
• African American leaders cried discrimination
o Insurance companies countered it was fair – their calculations were purely mechanical, and they merely used the data they had available
 In fact they are technically correct – it is impartial data usage, but discriminatory data selection; and thus a discriminatory final outcome
• In the 1860’s, the U.S. Government agreed with this objection

“JULIUS CHAPPELLE BILL” banned insurance programs based on ethnicity

Question 6

What was the disagreement in smoothing of insurance company’s financials?

Answer 6

• A core issue with this was seen in the setting of dividends:
o Logically, a year’s dividends should be based on the following yearly data:
 Number of policy holder deaths
 Company’s expenses
 Investment returns
o However, insurance companies set this based on the average of the previous few years instead of just that year’s data
 This reduced risk/uncertainty for the administration of the business
• Policyholders and legislators saw this as unfair
o Dividends aren’t directly correlated to profit
 Saw this as yielding corruption and sued
• This reflects two notions of capitalism
o “Wall Street Capitalism”
 Customer absorbs all risk of the market
 Managers just play a fair game
o “Corporated Capitalism”
 Should aim to maximize the benefit for the collective group of policy holders