When Growth Looks Healthy but People Still Feel the Squeeze
Life Essays
economy
jobs
AI
tech-and-society
2026-02-27 700 words

Lately I keep returning to one uncomfortable sentence: the economy is growing, but that growth feels less and less connected to ordinary people.

On paper, many numbers still look solid. In real life, people around us are still worried about hiring freezes, slower salary growth, and shrinking role demand. Both are true at the same time.

That contradiction is the point.


1. The Old Model Is Breaking

For years, we used a simple mental model:

That model worked well enough in a labor-heavy growth cycle. But now we are seeing a different pattern: companies can keep improving output without adding headcount, and sometimes by reducing it.

As engineers, we would call this a system rewrite, not a short-term incident.

In the old world, layoffs often looked temporary. In the new world, many roles are not paused; they are removed from the design.


2. Macro Wins, Micro Pressure

One reason this feels confusing is that we are looking at different metrics.

Macro indicators are like average throughput: they can look strong and stable. Individual experience is closer to P95/P99 latency: if the long tail gets worse, users still feel pain.

Translated to daily life:

A system can be objectively faster while more users report a worse experience.


3. Why This Time Feels Different

We used to replace human effort in layers:

  1. physical work with machines
  2. memory work with search and databases
  3. parts of reasoning with software

Now these layers are being merged into one production flow with AI + automation + robotics. When that stack works end to end, replacement speed is no longer linear.

A process that once required a full team can now run with:

From a company perspective, this is efficiency. From a worker perspective, this can be sudden displacement.

Both statements can be true.


4. Layoffs Are Becoming a Standard Tool

Another shift is cultural, not just technical.

In many organizations, layoffs are no longer treated as a last-resort emergency response. They are increasingly used as a routine management instrument to protect margins and improve “revenue per employee.”

That changes the game:

The irony is obvious in tech: we build tools to increase efficiency, then those same efficiency gains reduce demand for the teams that shipped them.

This is not about blaming workers for “not working hard enough.” It is about incentive structures and where value is captured after productivity increases.


5. What I Think Developers Should Do

I do not think the right response is panic, and I do not think denial helps either. We need a better model and practical strategy.

Three things feel clear:

First, waiting for a generic recovery to restore all old roles is risky. Some jobs are cyclical; many are now structurally replaced.

Second, coding skill is still essential, but “only coding” is becoming baseline. Higher resilience comes from combining engineering with business context, system design, process ownership, and communication across functions.

Third, career planning should move from job-title thinking to capability-stack thinking. Titles can disappear. Real problems do not.

If your work is close to core business pain points and hard to abstract away, your replacement risk goes down.


6. Final Note

I am not anti-tech, and I am not anti-AI. I work in this field and I benefit from these tools every day.

But technology does not distribute gains by itself. It optimizes what can be optimized. Whether that optimization improves life broadly or concentrates value in reports and valuations depends on policy, governance, and organizational choices.

That is why the “jobless growth” discussion matters.

The hard question is not whether development should continue. The hard question is: in an era of growth with fewer jobs, how do ordinary people still build a stable and dignified life?

This is where the real conversation starts.