Social Engineering Series, Part 7

The dose: the sources behind the cards.

A ten-slide series on the fourth move in the playbook: one small step at a time, so the next one always feels normal. This page has the studies, the court records, and the reporting behind every card, with dates, links, and what each one cannot tell you, plus a reading check you can try yourself.

Published October 15, 2026 · Emily Adams · Checked against the original studies, court records, and the reporting

The Cards

The full text of every card, in order, with the card itself under each one. Select a card to open it full size.

Card 01 of 09
Nobody asks you to accept the end point. Just the next step.
  • That is the fourth move. Go one small step at a time, so each step makes the next one feel normal and the first shock wears off. No single step looks like the big one.
SCOPE Every example ahead is from 2025 and 2026, with dates. The research on small steps is mixed, and the cards say where.
All sourcesCard 1
Card 02 of 09
The move, in three steps.
  • It works the same way in a war, a subscription, and a screen.
  • Start small. A first step that is easy to accept, with a reason attached.
  • Add a little. Each new step is only slightly bigger than the one you already accepted.
  • Let the baseline move. What shocked you last year is the starting point this year.
No single step looks like the big one. Add them up.
What these cards cannot tell youCard 2
Card 03 of 09
In the Lab. When blue dots became rare, people started calling purple dots blue. The line moved with what they had been seeing.
  • The same shift happened with threatening faces and unethical proposals, even when people were warned and paid to stay consistent.
  • In a lab advice game, self-serving exaggeration grew step by step, and the brain’s response to it faded as it grew.
  • Not always: in bribery experiments, people took a big bribe more readily when it was offered outright than after small ones.
SCOPE Seven studies of 21 to 91 Harvard students, 2018. Dots are not politics; that jump is ours. The exaggeration study had 55 adults, 25 of them scanned.
Sources and linksCard 3
Card 04 of 09
One boat, then an armed conflict.
  • The strikes in the Caribbean and the Pacific, step by step, in the administration’s words.
  • One boat. The president announces 11 killed. The next day: “it won’t stop with just this strike.” Sep 2025
  • A notice to Congress says the U.S. is in a “non-international armed conflict” with cartels. Oct 2025
  • Strikes reach the Pacific. An aircraft carrier arrives; about 12,000 troops are in the region. Nov 2025
  • After a report of a second strike on survivors of the first boat, the White House confirms it and calls it lawful. Dec 2025
  • An oil tanker is seized and a blockade ordered. Then a “large-scale strike” on Caracas and the capture of Maduro. Jan 2026
  • At least 235 people killed in 70 strikes since September 2025, the AP counts. Oct 2026
Each step was announced and defended. Add them up. Counts come from the administration’s own announcements. No court has ruled on whether the strikes are lawful.
Sources and linksCard 4
Card 05 of 09
One city, then the next.
  • National Guard and military deployments to U.S. cities, step by step.
  • Los Angeles: 2,000 Guard troops, then 2,000 more two days later, plus about 700 Marines. Jun 2025
  • Washington, D.C.: the Guard is called up “until I determine” order is restored, and other states send troops. Aug 2025
  • An executive order sets up a “standing National Guard quick reaction force.” Aug 2025
  • Memphis, Portland, Chicago, some with troops from other states. Courts block Portland and Chicago. Sep to Oct 2025
  • Troops leave three cities, which they may return to “in a much different and stronger form.” Dec 2025
  • The D.C. deployment is extended to Inauguration Day 2029. About 3,000 troops are there in October. Jul 2026
Courts stopped some steps. The baseline still moved. The administration’s own phrase for the plan, as quoted by a federal court: “city by city.”
Sources and linksCard 5
Card 06 of 09
At the Store. Netflix’s standard plan: $15.49, then $17.99 in January 2025, then $19.99 in 2026. “We will occasionally ask our members to pay a little more.”
  • Drip pricing: a low price first, then fees screen by screen. A federal rule against it for tickets and rentals began in May 2025.
  • April 2026: StubHub agreed to refund $10 million over fees shown across three screens. It admitted no wrongdoing.
  • Shrinkflation: a smaller package, same price. The government counts it as a price increase; shoppers notice it far less.
SCOPE Netflix shareholder letters, 2025 and 2026. Each price step was announced openly; the point is how small steps add up, not that anyone hid them.
Sources and linksCard 6
Card 07 of 09
At Home. Parents saying their child under 2 watches YouTube: 45% in 2020, 62% in 2025.
  • Daily screen time: 1 hour 3 minutes under age 2, 2 hours 8 minutes at 2 to 4, 3 hours 28 minutes at 5 to 8.
  • Own smartphone: 29% at ages 8 to 10, roughly six in ten at 11 to 12.
  • The administration’s May 2026 health advisory: exposure “often begins before a child’s first birthday and increases as children age.”
SCOPE Surveys of different children at different ages, 2020 to 2025, not one child followed over time. They show a staircase by age, not a cause.
Sources and linksCard 7
Card 08 of 09
What helps: write down your line.
  • The line moves with what you have recently seen. So set it before the next step, not after.
  • Write down where your line is now, and what would cross it. Read it again in six months.
  • Compare to the start, not to last month. Ask whether you would have said yes if this had been step one.
  • Add it up. Total the fees, the minutes, the troops, the strikes, then look at the total.
  • Notice when the only reason for a step is “we already did the last one.”
The question to keep: if this had been step one, would I have said yes?
SCOPE Levari et al., 2018: warning people, and paying them to stay consistent, did not stop the line from moving. That is why it goes on paper. These are habits, not tested fixes.
Sources and linksCard 8
Card 09 of 09
Your kids get it one notch at a time.
  • Screen time, apps, and what counts as normal at school all creep. Give them a way to notice.
  • “If someone had asked you for all of this on day one, would you have said yes?”
  • “What felt like a lot last year? What feels normal now?”
  • “You can go back a step. Normal is not the same as okay.”
  • “We set our limits before the app sets them for us.”
Set the line together, and write it down.
What these cards cannot tell youCard 9
Closing card
Next: who even knows anymore.
  • Part eight is exhaustion: so many claims, reversals, and fakes that people give up on knowing what is true. Every source on these cards is linked, with its limits.
Closing card

The Sources

Grouped by card. Each source has a summary of what it says, a scope line saying how far it reaches, and a link to the original.

Card 3

In the Lab

On card 3

Prevalence-Induced Concept Change in Human Judgment

Levari and colleagues, 2018 · Science, 360(6396), 1465 to 1467

Seven studies with Harvard students. People judged dots as blue or not blue, faces as threatening or not, and research proposals as ethical or not. When blue dots, threatening faces, or unethical proposals became rare, people widened the category: purple dots became blue, neutral faces threatening, ambiguous proposals unethical. The shift held when people were warned it would happen and when they were paid to stay consistent, and it worked the same whether the change was gradual or sudden. When blue dots became more common, the category shrank.

Scope  Small student samples, 21 to 91 per study, with artificial tasks. The paper is about how judgments track what people have recently seen. Applying it to politics or prices is my inference, not the paper’s finding.
On card 3

The Brain Adapts to Dishonesty

Garrett, Lazzaro, Ariely and Sharot, 2016 · Nature Neuroscience, 19(12), 1727 to 1732

Adults advised a partner about how much money was in a jar of pennies. When exaggerating helped the advisor, the exaggeration grew from trial to trial. In the 25 people scanned, the amygdala’s response to their own self-serving dishonesty faded over time, and the size of the fade predicted the size of the next increase.

Scope  One lab study in London: 55 adults in the main experiment, 25 of them scanned, and a second experiment with 25. Small stakes, no chance of being caught. It is about exaggerating a number, not serious wrongdoing.
On card 3

The Road to Bribery and Corruption: Slippery Slope or Steep Cliff?

Köbis, van Prooijen, Righetti and Van Lange, 2017 · Psychological Science, 28(3), 297 to 306

Four experiments with a bribery game found the opposite of a slippery slope for serious corruption: people were more likely to take a large bribe when it was offered outright than after first taking small ones. The authors describe the route to corruption as “a steep cliff rather than a slippery slope.”

Scope  I read the abstract, not the full paper, so no numbers are printed. It is on the card because the evidence on small steps is mixed and the card should say so.
Background context

Compliance Without Pressure: The Foot-in-the-Door Technique

Freedman and Fraser, 1966 · Journal of Personality and Social Psychology, 4(2), 195 to 202

The classic small-step study. Homeowners first asked to do something small were far more likely to agree later to a very large “Drive Carefully” sign in their yard: 76.0% against 16.7% of those asked cold.

Scope  Later reviews find the average effect is small, and a 2013 direct replication in Poland and Ukraine found no significant effect. That is why it is background here, not on the card.
Background context

Rapidly Declining Remarkability of Temperature Anomalies May Obscure Public Perception of Climate Change

Moore, Obradovich, Lehner and Baylis, 2019 · Proceedings of the National Academy of Sciences, 116(11), 4905 to 4910

In 2.18 billion posts on Twitter, what people treated as remarkable weather was set by the weather of roughly 2 to 8 years earlier. The authors call it evidence of the “boiling frog” effect, as a metaphor.

Scope  Twitter users are not a representative sample, and the study measures how much people post about weather, not what they believe.
Card 4

One Boat, Then an Armed Conflict

On card 4

The First Strike, September 2, 2025

Associated Press and CBS News, 2025 · Associated Press, September 3, 2025, as carried by ABC7; CBS News, September 3, 2025

The president announced a strike on a boat in the southern Caribbean, said 11 people were killed, and said the boat was operated by Tren de Aragua. AP noted the video did not show how many people were aboard or any clear drug stash. The next day Defense Secretary Pete Hegseth said on Fox and Friends that “it won’t stop with just this strike.”

Scope  The administration has offered little evidence about who was aboard the boats, and the card does not say what any boat carried.
On card 4

The Notice to Congress: “Non-International Armed Conflict”

NBC News and CBS News, 2025 · NBC News and CBS News, October 2, 2025

A notice to Congress said the president had determined that designated cartels are non-state armed groups whose actions amount to an armed attack, that “the United States is in a non-international armed conflict” with them, and that those killed are unlawful combatants. A White House spokesperson said the president “acted in line with the law of armed conflict.”

Scope  The wording is as quoted by NBC and CBS; I could not open the notice itself.
On card 4

A Timeline of U.S. Military Escalation Against Venezuela

Associated Press, 2025 and 2026 · Associated Press, January 3, 2026, as carried by PBS NewsHour

The AP timeline runs from the first strike to strikes in the Eastern Pacific from October 21, 2025; the aircraft carrier Gerald R. Ford ordered in on October 24 and arriving November 16, with about 12,000 troops in the region; an oil tanker seized December 10; a blockade of “sanctioned oil tankers” ordered December 16; and on January 3, 2026, a “large-scale strike” across Caracas and the capture of Nicolás Maduro. By then, AP counted 35 known boat strikes and at least 115 people killed, per the administration’s announcements.

Scope  Counts are from the administration’s own announcements, with almost no independent verification.
On card 4

White House Confirms Second September 2 Strike

NBC News, 2025 · NBC News, December 1, 2025; The Washington Post, November 28, 2025

After The Washington Post reported, citing anonymous sources, that a second strike killed two survivors of the first boat, the White House confirmed the second strike: “Secretary Hegseth authorized Adm. Bradley to conduct these kinetic strikes,” and said the admiral acted within his authority and the law. The Pentagon called the Post’s account of the order “completely false.”

Scope  What order was given is disputed and rests on anonymous sources; the card says only that the White House confirmed the strike and called it lawful.
On card 4

Strikes Continue: At Least 235 Killed in 70 Strikes

Associated Press, 2026 · Associated Press, October 5, 2026, as carried by CBS News

A strike on October 4, 2026 killed 4 people, bringing AP’s count to at least 235 people killed in 70 strikes since September 2025. Along the way the count was at least 37 (October 2025), 61 (late October), 95 (December), and 199 (May 2026). Congress voted down limits on the strikes by narrow margins, including a 50 to 50 Senate vote broken by the vice president in January 2026.

Scope  Other trackers count higher; I used AP’s count. No court has ruled on whether the strikes are lawful, and the card does not call them lawful or unlawful.
Card 5

One City, Then the Next

On card 5

Los Angeles and Washington, D.C.: The First Orders

The White House, 2025 · The White House, presidential memoranda of June 7 and August 11, 2025, and Executive Order 14339 of August 25, 2025

June 7, 2025: a memorandum federalized “at least 2,000” California Guard for 60 days over the governor’s objection; two days later came 2,000 more and about 700 Marines. August 11: the D.C. Guard was mobilized “until I determine that conditions of law and order have been restored,” and other states were invited to send troops. August 25: an executive order called for trained civil-disturbance units in every state’s Guard and a “standing National Guard quick reaction force.”

Scope  The link is the D.C. memorandum; the June 7 memorandum and the August 25 order are on the same White House archive. Troop numbers are as published at each step.
On card 5

Portland, Chicago, and the Courts

NPR and the Supreme Court, 2025 · NBC News, October 2025; Supreme Court of the United States, Trump v. Illinois, December 23, 2025; NPR, December 23, 2025

Guard troops were federalized for Portland (200) and Chicago (at least 300) in September and October 2025, with troops from California and Texas sent across state lines. Federal trial courts blocked both. On December 23, 2025, the Supreme Court, 6 to 3, declined to let the Chicago deployment go ahead while the case continued.

Scope  The Supreme Court’s order was interim and limited to the law used in Illinois. It did not decide the D.C. deployment, which rests on different law, and courts also sided with the administration at several steps.
On card 5

Withdrawal From Three Cities

CNN, 2025 · CNN, December 31, 2025, as carried by KEYT

The president announced the Guard would leave Chicago, Los Angeles, and Portland, adding that troops could return “perhaps in a much different and stronger form.” The withdrawals were completed in January 2026.

Scope  This is a step back, and the card says so: courts stopped some steps.
On card 5

Washington, D.C., Extended to 2029

NPR and AP, 2026 · NPR, July 14, 2026; Associated Press, August 4 and October 2, 2026

The Pentagon confirmed the D.C. deployment would continue to Inauguration Day 2029, with nearly 5,000 troops during the summer of 2026. In August the Pentagon estimated about $1.4 billion to keep an average of 2,500 there through January 2029. AP reported about 3,000 troops in October 2026.

Scope  Troop numbers changed through the year; each one is printed with its date and outlet.
Background context

“City by City”

Federal court record, 2025 · As quoted in a federal court opinion in the Illinois case, from an administration statement of October 6, 2025

The administration’s own phrase for the plan, “city by city,” was quoted by a federal court in the Illinois case. The link is the Supreme Court’s docket for that case, which includes the lower-court record.

Scope  I read the phrase as quoted by the court, not in the original statement.
Card 6

At the Store

On card 6

Shareholder Letters, January 2025 and April 2026

Netflix, 2025 and 2026 · Netflix, Exhibit 99.1 to Forms 8-K filed with the SEC, January 21, 2025 and April 16, 2026

In January 2025 Netflix raised its U.S. standard plan from $15.49 to $17.99 a month, telling shareholders “we will occasionally ask our members to pay a little more.” By 2026 the standard plan was $19.99, and the plan with ads had gone from $6.99 to $7.99 to $8.99.

Scope  Each step was announced openly. The card is about how small steps add up, not about anything hidden. I could not confirm the month of the 2026 increase, so the card says 2026.
On card 6

The Junk Fees Rule, and StubHub

Federal Trade Commission, 2025 and 2026 · FTC press releases, May 2025 and April 2026

A federal rule requiring the full price up front for live-event tickets and short-term lodging took effect on May 12, 2025. In April 2026 StubHub agreed to refund $10 million to settle FTC charges that it revealed fees across three screens at checkout.

Scope  These are the FTC’s allegations; StubHub did not admit wrongdoing.
On card 6

Shrinkflation and How Shoppers Respond

Government Accountability Office, 2025 · GAO-25-107451, July 31, 2025; Bureau of Labor Statistics

The Bureau of Labor Statistics counts a smaller package at the same price as a price increase. GAO reviewed the research and found studies in which shoppers were about two to five times more responsive to a price increase than to an equal cut in package size.

Scope  I did not print specific product examples, because the ones I found came from secondary sources.
Card 7

At Home

On card 7

How Parents Manage Screen Time for Kids

Pew Research Center, 2025 · Pew Research Center, October 8, 2025

A survey of 3,054 parents of children 12 and under, May 2025. Among parents of a child under 2, the share who say the child watches YouTube rose from 45% in 2020 to 62% in 2025. 29% of parents of 8- to 10-year-olds and roughly six in ten parents of 11- and 12-year-olds say the child has a smartphone.

Scope  Parent reports. The 2020 survey asked about children 11 and younger; Pew says results are largely unchanged when restricted the same way.
On card 7

The 2025 Common Sense Census: Media Use by Kids Zero to Eight

Common Sense Media, 2025 · Common Sense Media, February 26, 2025

A nationally representative survey of 1,578 parents, August 2024. Average daily screen time: 1 hour 3 minutes under age 2, 2 hours 8 minutes at ages 2 to 4, and 3 hours 28 minutes at 5 to 8. Forty percent of children have a tablet by age 2.

Scope  A survey of different children at different ages, not one child followed over time.
On card 7

Surgeon General’s Warning on the Harms of Screen Use

U.S. Department of Health and Human Services, 2026 · HHS advisory, May 20, 2026

The advisory says exposure “often begins before a child’s first birthday and increases as children age.” Its section for families suggests limits that could be none under 18 months, under one hour a day under age 6, and two hours a day for ages 6 to 18.

Scope  An advisory, not a regulation. There was no confirmed Surgeon General when it was issued; news reports say HHS officials developed it.
Background context

Does the Feed Push People Toward Extremes?

Hosseinmardi and colleagues, 2024; Chen and colleagues, 2023 · Proceedings of the National Academy of Sciences, 2024; Science Advances, 2023

The strongest studies of YouTube found little evidence that the recommendation algorithm pushes adult users step by step toward extreme content; most extreme viewing came from people who already sought it out.

Scope  These studies are of adults. I found no 2025 or 2026 audit of what feeds show teen accounts, so the home card is about time and habits, not content.

Check Your Reading

Read the article below, then try the six questions. Each one uses a different reading skill from part two, in rough order of difficulty. Write your answer down before you open the strong answer.

Read this first

Supreme Court rules against Trump’s emergency tariffs, but leaves key questions unanswered

Kent Jones, Babson College · The Conversation · February 20, 2026 · about 970 words · free to read
  1. DecodeRead the words and know what they mean here. · Level 1

    Paragraph 10 says the majority treated tariff revenue as part of Congress’s “taxing power.” What does that mean here?

    See a strong answer

    The Constitution gives the power to tax to Congress. The Court treated tariffs as a kind of tax, so a president needs authority from Congress to impose them.

    Paragraphs 9 and 10

  2. LocateFind a specific piece of information in the text. · Level 1 to 2

    What was the vote in the case, and on what date was it decided?

    See a strong answer

    6 to 3, on February 20, 2026.

    Paragraph 2

  3. RestateSay what it means in your own words. · Level 2

    In one or two sentences of your own, what did the Court decide, and what did it leave open?

    See a strong answer

    The Court ruled that the 1977 emergency powers law does not let the president impose tariffs, which removes many of his broadest tariffs. It left open other tariff laws, what happens to deals built on the old tariffs, and whether companies get refunds.

    Paragraphs 1, 2, 12 and 19

  4. InferWork out what the text implies but does not say. · Level 3

    Paragraphs 12, 15 and 17 say the ruling does not reach Section 232 or Section 301, and that Section 232 needs a separate action for each industry. What would you expect to happen next?

    See a strong answer

    Tariffs coming back piece by piece under other laws, one industry or one country at a time: a new series of steps rather than one big one. The article does not predict this; you have to put the paragraphs together. It is the move on this part’s cards.

    Paragraphs 12, 15 and 17

  5. CompareSet two ideas, claims, or sources side by side. · Level 3

    According to paragraphs 7 and 8, how is the emergency powers law different from other tariff laws?

    See a strong answer

    The emergency law allows things like embargoes and asset freezes but does not mention tariffs. Other tariff laws spell out how tariffs may be imposed.

    Paragraphs 7 and 8

  6. EvaluateJudge how far to trust it, and what to check. · Level 4 and up

    Paragraph 5 says the fentanyl tariffs included 25% on some goods from China, and paragraph 2 says the use of the law was “unconstitutional.” Check both against the sources the article links. What do you find?

    See a strong answer

    The White House fact sheet the article links says 25% on Canada and Mexico but 10% on China, later raised to 20%. And the Court’s holding was that the law “does not authorize the President to impose tariffs,” a ruling about what the law allows, not a finding that anything was unconstitutional. That difference matters: Congress could pass a law that allows such tariffs. Even a careful explainer by an expert can get details wrong, which is why you check the source.

    Paragraphs 2 and 5, with the White House fact sheet and the Court’s opinion

Connect it to the cards: the tariffs went up and down in steps, and the next steps, under other laws, came one at a time. Download the printable worksheet (PDF), with room to write and an answer key. Want more practice? Play Between the Lines: read a current news article, answer twelve questions, and see your reading level.

What These Cards Cannot Tell You

The research on small steps is mixed. Some studies find that small steps lead to bigger ones; a careful bribery study found the opposite, a steep cliff rather than a slope. The card shows both.
Dots are not politics. The strongest lab finding is about how people judge dots, faces, and proposals after the mix changes. Applying it to wars, prices, and screens is my inference.
A sequence is not a plan. The cards show steps in order, with dates. They do not claim anyone planned the whole sequence in advance, and they report each step’s stated reason.
No court has ruled on the boat strikes. The cards do not call them lawful or unlawful. The counts come from the administration’s own announcements.
Courts stopped some steps. Trial courts blocked deployments in several cities, the Supreme Court kept the Chicago deployment blocked, and troops left three cities. Appeals courts also sided with the administration at several steps. The card says some steps were stopped.
The price steps were announced. Netflix told its customers and shareholders. The point is how small steps add up, not that anyone hid them.
The home numbers are a staircase by age, not one child’s path. They come from surveys of different children at different ages.
The lines for kids are mine. The four sentences on card 9 are things I would say at my own table. They are not findings.

Why this one is easy to miss.

The first two moves at least happen in a moment you can point to: a joke, a repeated line. This one has no moment. Each step is small enough to accept, and by the time the total is visible, the starting point has moved.

What stayed with me from the research is that warning people did not stop it. In the dot studies, people who were told the mix would change, and even people paid to stay consistent, still moved their line. Knowing about the move is not enough. You need something outside your own memory to compare against.

That is why the advice on card 8 is so plain: write the line down, and compare to the first step, not the last one. It is the same thing I want my kids to do with their screens, and the same thing I am trying to do with the news.