Social assistance says it wants people to become more active and independent. Work a few hours. Take some training. Try self-employment. Share accommodation. Form relationships. Accept help from family. Save a little money. Find cheaper housing. Travel to improve opportunity. Participate in the community. Improve your circumstances.
Then try doing any of those things.
Earn some money and the system wants to know about it—not simply because it is interested in your progress, but because some of that income may reduce your benefits. Receive an honorarium and it has to be classified. Get help from family and the system may want to know whether it is a gift, income or continuing support. Change your living arrangements and questions arise about who lives with you, what they contribute and the nature of the relationship. Accumulate resources and asset rules can become relevant.
And if you live in rent-geared-to-income housing, another system can be watching some of the same changes from the opposite direction. Social assistance may ask whether your additional income means it can pay you less while the housing system asks whether the same additional income means it can charge you more.
The message is remarkably contradictory:
Please improve your circumstances. And when you do, report back so that we can calculate what we can take away.
That is not merely a communications problem or an unfortunate amount of paperwork. It changes behaviour.
Over the years I have met three people receiving social assistance who had arrived at what may be the safest possible strategy for living within such a system. Two lived in rent-geared-to-income housing and the third in a rooming house. They earned nothing. They had virtually no financial activity. They accepted little help that might complicate their circumstances. Their living arrangements did not change. They did not travel outside the province. They accumulated almost nothing.
They lived remarkably inert lives. Their inert lives were purposeful.
And, viewed purely through the logic of the system, their behaviour made perfect sense.
There was almost nothing to report because almost nothing happened. There was little for social assistance to reassess because there was little to change. For the two in RGI housing, there was little new income with which to recalculate the rent. Their poverty remained severe, but their administrative lives became comparatively stable.
They had discovered something about social assistance that our rhetoric about “incentives” tends to miss.
In a system that responds to improvement by reassessing what it gives you and what it charges you, doing nothing can be a rational form of self-protection.
We reward inactivity with certainty
Consider two people receiving the same inadequate social-assistance income.
The first tries to improve his situation. He takes occasional shifts. He accepts a small honorarium for community work. His sister helps him through a difficult month. He considers sharing an apartment. Perhaps he takes a course. Perhaps he starts a tiny business. He is doing precisely what governments say they want people receiving social assistance to do.
But each movement creates a transaction with government.
What did you earn? When did you earn it? Was it deposited in you bank account? Is the honorarium exempt? Was the money from your sister a gift? How much was it? Will she give you more? Who are you living with? Are you sharing expenses? What is the nature of that relationship? Has your rent changed? Are you still eligible for the same amount?
And then comes the crucial second step.
The system doesn’t collect all that information merely to admire initiative.
It collects the information because the answer can change what people gets.
The second recipient does none of those things. No employment. No honoraria. No new household arrangements. No meaningful savings. No travel. No financial help that needs explaining. Nothing changes.
Which recipient has the quieter file?
The second.
Which recipient has fewer reporting obligations?
The second.
Which has fewer opportunities for a misunderstanding, recalculation, overpayment or interruption?
The second.
Which person can predict next month’s income with greater confidence?
Quite possibly the second.
And if both live in RGI housing, the contrast can become still sharper. The active recipient may have one system recalculating benefits downward while another recalculates rent upward. The inactive recipient provides neither system with much to recalculate.
We have somehow built an income-security system in which nothing happening can become the safest thing that can happen.
We have misunderstood the word “incentive”
Governments usually discuss welfare incentives as though recipients face a relatively simple financial calculation: if someone earns another dollar, how much of that dollar do we allow them to keep?
Hence earnings exemptions. Hence discussions of benefit-reduction rates. Hence the familiar policy objective of “making work pay.”
Those things are crucial. But they capture only part of the calculation made by someone who has virtually no financial margin for error.
The real calculation can be closer to this:
If I earn this money, what will happen to my assistance? What will happen to my rent? Will another benefit change? Will I have to produce documents? Will the amounts be calculated correctly? Will there be an overpayment? Will next month’s payment arrive as expected? And after everybody has recalculated everything, how much better off will I actually be?
For a middle-income household, a bureaucratic mistake is irritating. For someone living on social assistance, an unexpectedly reduced payment can mean not paying the phone bill, not buying groceries or not making the rent.
Certainty therefore has economic value.
And when someone has no buffer, avoiding risk can be worth more than pursuing a modest gain.
This is where very low social-assistance rates can perversely work against the independence policymakers say they want. When every dollar of the existing payment is needed for survival, the rational priority can become protecting the payment rather than experimenting with ways of eventually escaping it.
The poorer we make someone, the less capacity that person has to take the risks involved in becoming less poor.
The system wants movement but penalizes it
This is the contradiction we rarely confront directly. We want recipients to work, but earnings can reduce benefits. We want people to build financial resilience, but needs-tested systems scrutinize resources.
We want families to help each other, but assistance from family can become something that has to be reported, classified and sometimes taken into account.
We want people to form stable relationships and share expenses, but household arrangements can become eligibility questions.
We want people to participate in their communities, but even a modest honorarium can create another reporting and classification exercise.
We want people to find affordable ways to live, including sharing accommodation, but changing household arrangements can trigger questions about costs, contributions and relationships.
And when someone lives in RGI housing, increased income can produce another consequence entirely: the person may simultaneously encounter a welfare system designed to pay less as income rises and a housing system designed to charge more.
Each rule can be explained individually.
Together they create a system.
And the system sends a message very different from the one contained in government brochures about independence:
Move forward and we recalculate it. Stay exactly where you are and we may leave you alone.
The supervision/surveillance paradox.
We have turned getting ahead into an eligibility event
Most people construct independence through a messy accumulation of small advantages.
They take an extra shift. Someone gives them a hand. They move in with somebody. They split costs. They save a little. They take a course. They try a sideline. A relative pays an unexpected bill. They build a relationship. They move. They fail at something and try something else.
We call this resourcefulness. For someone receiving deeply needs-tested assistance, most of the same behaviours become reportable events.
That should trouble us. The system says it wants recipients to acquire the scaffolding of independence while simultaneously inspecting each piece of scaffolding to determine whether it justifies withdrawing some assistance.
In ordinary life, family help is a resource. Inside a needs-tested system, it can become a question.
In ordinary life, a new relationship can increase resilience. Inside the system, it can become a file review.
In ordinary life, additional earnings are evidence that someone is getting ahead. Inside the system, they can become an opportunity to reduce benefits and increase RGI rent.
The behaviour has not changed. Its administrative meaning has.
And then we blame the person for dependency
This is where the contradiction becomes particularly unfair.
Imagine someone observing one of the three people I encountered. No job. No travel. No savings. Few changing circumstances. An extraordinarily narrow life organized around maintaining a very small but predictable income.
It would be easy to point to that person and say:
“See? Welfare creates dependency.”
But that explanation may have the causation backwards. Perhaps the person has learned how to minimize exposure to a system that makes change consequential.
Perhaps inactivity is not evidence that incentives have failed. Perhaps inactivity is the response to the incentives we actually created.
If that is true, the traditional policy response—more conditions, more monitoring, more reporting, more supervision—can make the problem worse. Every additional rule increases the number of things that can go wrong when a recipient does something.
We then supervise people more intensely because they are insufficiently active while making activity increasingly consequential to their entitlement.
That is not an incentive system. It is a trap.
The three people who understood the rules
I don’t suggest that the three people I encountered represent social-assistance recipients generally. They plainly do not. Most people’s lives are much messier. People work, lose jobs, form and leave relationships, receive family help, become sick, recover, move, study and deal with changing expenses.
That is precisely why these three unusually inert lives stayed with me. They represented something close to a pure response to the architecture. They had reduced the number of variables. No earnings to report. No honoraria to explain. No changing household finances. No travel. Little outside assistance.
Almost nothing that could cause one bureaucracy to ask whether it should give them less or another to ask whether it should charge them more.
They were still poor. They had simply made their poverty administratively uncomplicated. And there is something profoundly wrong with a system when that can be a rational achievement.
If we want movement, make movement safe
The answer is not to stop caring about income or household circumstances. A needs-tested program cannot operate without some rules.
But we can ask a different design question. Instead of continually asking how to make recipients behave differently, ask:
What happens to a recipient when they behave exactly as we say we want?
Take every desirable activity—earning, training, saving, accepting family help, sharing accommodation, forming relationships, participating in the community—and run it through the entire income-security system.
Does the person end up meaningfully better off? How many agencies need to know? How many calculations change? How much additional rent becomes payable? How much benefit disappears? How much documentation is required? How much risk does the person assume? And, at the end of it all, was becoming more independent actually worth it?
That should be the real test of welfare reform.
Good policy would create much larger spaces in which ordinary life could occur without constantly recalculating poverty: substantial earnings exemptions, reasonable treatment of modest gifts and honoraria, sensible asset rules, portable benefits, transition periods, fewer cliffs and less intrusive treatment of ordinary relationships. RGI housing and social assistance should also be examined together rather than pretending their simultaneous clawbacks occur in separate universes.
Most importantly, we should recognize that people respond not to the incentives policymakers announce but to the incentives systems actually create.
For decades, social assistance has told recipients that the objective is independence. But independence consists of doing things: working, saving, sharing, moving, accepting help, forming relationships and taking chances.
We have constructed a system that watches those things happen because each may provide a reason to reconsider what the person receives. Then we wonder why some people become extraordinarily careful about doing anything at all.
The three people I met had not failed to understand the system. They may have understood it perfectly.
They had discovered the one lifestyle that caused almost nobody to ask them for anything, almost nobody to recalculate anything and almost nobody to look for something more to take away.
They stayed poor. They stayed still. And the system finally left them alone.