Cheaper AI, More Informality? A Dual Labor Market Model for Developing Economies

The Great Robot Price Drop: A Story of Two Worlds

Imagine the economy as a giant, bustling kitchen. In this kitchen, there are two distinct zones. One is the "Formal Kitchen," where chefs wear crisp uniforms, follow strict health codes, pay taxes, and get paid a steady wage. The other is the "Informal Kitchen," a chaotic but lively street stall next door where cooks work without uniforms, no one pays taxes, and the pay is whatever the customer hands over that day. In many parts of the world, especially in Latin America, the Informal Kitchen is huge—sometimes bigger than the Formal one.

Now, imagine a new, super-smart robot arm arrives. It can chop vegetables, stir sauces, and even write recipes. The big question economists are asking is: What happens when this robot arm gets cheaper? Does it help the Formal Kitchen hire more chefs and make better food? Or does it kick the human chefs out, forcing them to flee to the Informal Kitchen to survive?

To answer this, we need to understand one simple idea: substitution. Think of it like a dance. If the robot and the human are dancing partners who need each other to move (complements), making the robot cheaper means they dance together more, and the human gets a better job. But if the robot and the human are doing the exact same dance moves (substitutes), making the robot cheaper means the boss fires the human to save money. This paper explores exactly what happens in our two-kitchen economy when the price of these "AI robots" drops, and why the answer depends entirely on whether the robots are partners or replacements.

The Paper's Big Discovery: It's All About the "Buddy System"

In this paper, Gabriel Montes-Rojas, Fernando Toledo, and Juan Manuel Rodriguez Repeti build a digital simulation of a developing economy (calibrated to look like a typical Latin American country) to see what happens when Artificial Intelligence (AI) gets cheaper. They don't just guess; they run a complex computer model that tracks how money, jobs, and prices move between the Formal and Informal sectors.

The authors find that the outcome isn't a single story. Instead, the economy splits into two very different realities depending on a specific "personality trait" of the AI: Is it a substitute or a complement?

Scenario 1: The "Replacement" Robot (Substitution)
Imagine the AI is a robot that does the exact same job as a human worker. In the paper's model, they call this the substitution regime (where the "elasticity of substitution" is greater than 1).

• What happens: When the price of this robot drops, the Formal Kitchen bosses say, "Great! Let's fire half our chefs and buy more robots."
• The Result: Formal jobs shrink. The fired workers don't disappear; they flood into the Informal Kitchen. The Informal sector acts like a giant safety net or a shock absorber, absorbing all the displaced workers.
• The Twist: Even though people lost formal jobs, the overall economy actually gets a little richer because the robots are so cheap. The Formal Kitchen produces more, and the extra money trickles down, keeping wages from crashing completely. But the Informal sector gets crowded and messy.

Scenario 2: The "Buddy" Robot (Complementarity)
Now, imagine the AI is a robot that is terrible at doing the job alone but is amazing at helping a human do it faster and better. In the model, this is the complementarity regime (where the elasticity is less than 1).

• What happens: When the robot gets cheaper, the Formal Kitchen bosses say, "Wow, with this cheap robot, my chefs can work twice as fast! Let's hire more chefs!"
• The Result: Formal jobs explode. The Formal Kitchen expands, wages go up, and everyone is happier. The Informal sector also grows, but not because people are being kicked out. Instead, the whole economy is booming, and the Informal Kitchen is busy serving the extra customers created by the new wealth.
• The Winner: This is the best-case scenario. The economy grows, formal jobs increase, and the informal sector expands as a healthy part of a larger pie, not as a dumping ground for the unemployed.

The "Buffer" and the "Switch"

The paper highlights a crucial mechanism: the Informal Sector as a Buffer. In the simulation, when the Formal sector shrinks (because of the Replacement Robot), the Informal sector instantly swallows the extra workers. It's like a sponge soaking up a spill. The authors show that this buffer is vital for keeping the economy from collapsing, even if it means more people are working without benefits.

However, the paper also warns that the "switch" between these two worlds isn't always obvious. The authors ran simulations showing that the tipping point isn't exactly where you might think. Even if robots are mostly replacements, the economy might still create more formal jobs if the demand for goods is strong enough. But if the robots are strong replacements, the Informal sector takes the hit.

What the Paper Rules Out and What It's Not

It's important to know what this paper doesn't say.

• It doesn't say AI will definitely destroy jobs. The paper explicitly rules out the idea that cheaper AI always leads to unemployment. It shows that if AI helps humans (complementarity), jobs can actually increase.
• It doesn't say the Informal sector is "bad" in a moral sense. The model treats the Informal sector as a necessary economic function that stabilizes the economy during shocks.
• It doesn't prove these outcomes in the real world yet. The authors are careful to state that these are simulation results. They built a mathematical model and ran it on a computer. They haven't measured the real world yet to see which scenario (substitution or complementarity) is actually happening right now in Latin America. They suggest that the answer depends on a specific number (the elasticity of substitution) that policymakers need to measure carefully.

The Takeaway for Policymakers

The paper concludes with a playful but serious warning for leaders. Because the outcome depends on whether AI is a "buddy" or a "replacement," the solution depends on knowing which one it is.

• If AI is a replacement, the government needs to build a stronger safety net to catch the people falling out of the Formal Kitchen.
• If AI is a buddy, the government should focus on training workers to be better partners for the robots, so the Formal Kitchen can grow even bigger.

The authors suggest that the most important thing a country can do is figure out the "personality" of the AI in their specific industries. If they get this wrong, they might try to protect workers from a threat that isn't there, or fail to protect them from a threat that is. In the end, the paper shows that the future of work isn't written in stone; it depends on how we choose to pair our human skills with our new digital tools.