The task
The two articles below take opposite positions on one question:
Should the City of Millbrook make its guaranteed-income pilot permanent, paying 400 low-income households $500 a month with no conditions?
Read both. Then analyze both positions to determine which one is better supported by evidence. Use specific evidence from both articles in your answer.
You are not being asked which side you agree with, and you will not be scored on which one you choose. You are being asked to judge the arguing — and either article can be argued for. Allow about 45 minutes.
Article 1
What $500 a Month Did
1For two years, the Millbrook Bridge Pilot gave 200 low-income households $500 a month, with no conditions on how they spent it. More than 2,300 households applied. We chose the 200 by lottery and followed another 200 applicants who were not chosen, so that we could compare the two groups.
2The results were clear. At the start, 26 percent of the people who received the money worked full time. After two years, 37 percent did. In the comparison group, full-time work barely moved, from 27 percent to 29 percent. With a little breathing room, people could turn down the worst shifts, fix the car that got them to work, and look for something better.
3The money also bought stability. At the start, 29 percent of recipients said they could cover a surprise $400 expense. After two years, 52 percent could. In the comparison group, the figure rose only from 27 to 33 percent. And the money went to necessities. Of the money spent with the program’s debit cards, the largest shares went to groceries (34 percent), housing and utilities (22 percent), and transportation (15 percent). Less than 1 percent went to alcohol or tobacco.
4Millbrook’s results match what other cities have found. In Stockton, California, a pilot gave 125 residents $500 a month for two years. Full-time employment among recipients rose from 28 percent to 40 percent, a far larger gain than in the group that did not receive the money.
5We sometimes speak about poor families as if they cannot be trusted with money. The pilot tested that belief, and the belief was wrong. Millbrook should keep what works.
Article 2
Good Intentions, Uncertain Results
1The Bridge Pilot was paid for by private donors, and their grant has ended. Making the program permanent at 400 households would cost the city $2.4 million a year in payments, plus about $300,000 to run it. That money would come from the general fund, the same budget that pays for our summer youth jobs program. That program costs $2.1 million and employs 1,100 teenagers.
2Before spending that much, we should look at the largest and most careful study of guaranteed income in this country. In 2024, a team of researchers reported the results of an experiment that gave 1,000 low-income people in Texas and Illinois $1,000 a month for three years. A comparison group of 2,000 people received $50 a month. The people who received $1,000 worked between one and two fewer hours a week than the comparison group, and their income from sources other than the payments fell by about $1,900 a year. Improvements in stress and mental health faded after the first year, and the study found no effect on physical health.
3Our own pilot’s results deserve a closer look. They come from surveys that the families filled out themselves, and 31 percent of the families who received the money did not complete the final survey. The pilot was run, and its results were reported, by the same organization that is now asking the city to fund it.
4The city already has programs that help families find steady work, such as job training, help with child care, and help filing for tax credits. If the council wants to spend another $2.7 million on families in need, it should put the money into programs like those.
5I do not doubt that $500 a month helped the families who received it. Any family would be helped by more money. The question for the council is whether this is the best use of $2.7 million a year, and the evidence so far does not show that it is.
Your 45 minutes
Stop here if you have not written yet. What follows names the evidence in both articles and shows where each one is weak. Read it before you write and you will not be practicing the thing the test measures — which is finding all of this yourself, in forty-five minutes, with nobody helping.
After the clock — what was actually on the table
What each article put in front of you
What $500 a Month Did
- A local experiment with a comparison group. Two hundred households received the money, chosen by lottery, and 200 applicants who were not chosen were followed. A comparison group (also called a control group) is the group that does not receive the program. It shows what would have happened anyway, so the difference between the two groups is what the program did.
- Local results, before and after, for both groups. Full-time work: 26 to 37 percent for recipients, 27 to 29 percent for the comparison group. Able to cover a $400 surprise: 29 to 52 percent, against 27 to 33 percent.
- A record of spending. Groceries 34 percent, housing and utilities 22 percent, transportation 15 percent, alcohol and tobacco under 1 percent.
- A second city with similar results. Stockton: 125 residents, $500 a month for two years, and full-time employment from 28 to 40 percent.
Good Intentions, Uncertain Results
- A cost, with something to compare it to. $2.4 million a year in payments plus $300,000 to run the program, set beside a summer jobs program that costs $2.1 million and employs 1,100 teenagers.
- The largest study in the country. 1,000 people receiving $1,000 a month for three years, 2,000 in a comparison group: one to two fewer hours of work a week, about $1,900 a year less in other income, and health gains that faded.
- A specific problem with the local results. The results are self-reported, and 31 percent of recipients did not complete the final survey.
- A question about who is reporting. The organization that ran and reported the pilot is the one asking for the money.
Where each one is thin
Neither article is the right answer. Each has evidence a careful reader can question. Naming those weak points, in specific words, about a specific sentence, is most of what the first scoring trait rewards.
What $500 a Month Did
- The groups are small. Two hundred households in each group, and 125 people in Stockton. Small groups can show large changes by chance. A careful reader should be cautious about results from groups this size.
- She does not mention missing surveys. Article 2 says 31 percent of recipients did not complete the final survey. If the people who stopped answering were doing worse, the results would look better than they really were. Article 1 says nothing about this.
- The debit-card figures show only what was bought with the card. If recipients could also withdraw cash, the card cannot show how the cash was spent. She reports the card figures as if they were all of the spending.
- She does not discuss cost. The program would cost $2.7 million a year. Article 1 never names a figure or says where the money would come from.
- Her organization ran the pilot and would run the permanent program. That makes her the person who knows it best. It also gives her a reason to see it in a good light.
Good Intentions, Uncertain Results
- The large study tested a different program. It paid $1,000 a month for three years, in Texas and Illinois. Millbrook’s program pays $500 a month. Mr. Kessler does not explain why results from a larger payment, in other places, should decide the question for Millbrook.
- He leaves out part of the study. The same research found that recipients spent more on basic needs. A writer who reports only the findings that support his side is choosing his evidence, and a careful reader notices.
- He does not answer the local employment result. Article 1 reports that full-time work rose for recipients and not for the comparison group. Article 2 questions how the results were collected but never says that this finding is wrong.
- His alternative has no evidence. He says money should go to job training and child care, but gives no results for those programs in Millbrook or anywhere else.
- The summer jobs comparison is a choice, not a fact about the pilot. Setting $2.4 million beside a program for teenagers makes the cost feel large. It does not show that the guaranteed income failed.
Two essays, either of which could score well
If both of those look like real essays to you, the set is working as intended. On the test you will have two articles built the same way, and no time to find a perfect answer, because there is not one. Choose the side you can support in forty-five minutes, and support it.
Check your own
The Extended Response is scored on three traits. Read your draft against them, one at a time — and then read How the essay is scored, which has the full rubric and two responses to the same task, one earning a 6 and one earning about a 2.
- Analysis of argument and evidence. Did you say which article is better supported, in a sentence, near the beginning? Did you quote or name specific evidence from both articles — not just the one you chose? Did you say what is wrong with the evidence on at least one side, rather than only what is right?
- Development and organization. Does each paragraph do one job? Does a reader who has not seen the articles know what they said? Did you get to the end, or did you run out of clock in the middle of a paragraph?
- Clarity and command of English. Read it aloud. Sentences that you stumble over reading are sentences a scorer stumbles over too. Check that every sentence has a subject and a verb and ends where it should.
One more, which is not on the rubric and costs you the most: did you argue about which side is right instead of which side is better supported? That essay can be beautifully written and still score low, because it answers a question nobody asked.