Case Study 2 — The 2008–2012 U.S. Housing Market Through Supply and Demand

The U.S. housing market between 2003 and 2012 is one of the most studied — and one of the most dramatic — episodes of supply and demand in the modern economic record. House prices rose rapidly from 2003 to 2006, then collapsed by roughly 30% on average between 2006 and 2012, then recovered slowly. Why? What does the supply-and-demand model say about it? And what does the model leave out?

This case study uses real data from the Case-Shiller national home price index (FRED series CSUSHPINSA) and several other publicly available series to walk through the episode. You can reproduce all the numbers yourself in FRED. The interpretation is mine; the data is the official record.

The price record

Here is the rough trajectory of the Case-Shiller national home price index, indexed to 100 in January 2000:

• January 2000: 100 (definition)
• January 2003: ~130 (modest growth)
• January 2006: ~185 (the peak — 85% above 2000)
• January 2009: ~150 (down 19% from peak)
• January 2012: ~135 (the trough — 27% below peak)
• January 2020: ~210 (full recovery and beyond)
• January 2024: ~315 (roughly 70% above the 2006 peak)

The 2003–2006 run-up was historically extreme. Real (inflation-adjusted) home prices had been roughly stable for most of the postwar period; the 2003–2006 increase was the largest in U.S. history. The 2006–2012 collapse was also historically extreme: never before had nationwide home prices fallen so far so fast in the postwar period. (We discussed in Chapter 2 how the financial models that priced mortgage-backed securities had assumed regional independence in housing markets — and how that assumption broke when the entire country experienced a downturn at once.)

Now let's use supply and demand to explain it.

The 2003–2006 run-up: a demand-driven story

What happened to demand for housing in the early 2000s? Several things, and they all pointed in the same direction: right.

Demand shifter 1 — Lower interest rates. The Federal Reserve cut the federal funds rate from over 6% in 2000 to 1% by mid-2003 in response to the dot-com recession. Mortgage rates followed: 30-year fixed mortgages went from about 8% in 2000 to about 5.5% in 2003 and stayed below 6% for years. Lower mortgage rates dramatically reduce the monthly cost of buying a home for the same price — making housing more affordable in monthly-payment terms — which shifts demand right.

Demand shifter 2 — Looser lending standards. Banks lowered their lending standards substantially during the 2003–2006 period, lending to borrowers with lower credit scores, lower down payments, less documentation, and higher debt-to-income ratios. The famous "NINJA loans" (No Income, No Job, no Assets) were the extreme version, but the pattern of relaxed standards extended throughout the lending market. This effectively expanded the pool of people who could buy homes, shifting demand right.

Demand shifter 3 — Speculation. As prices rose, more people bought homes expecting prices to keep rising, hoping to flip them for profit. Some bought multiple houses on speculation. This is the expectations shifter: when buyers expected continuing price increases, current demand shifted right.

Demand shifter 4 — Demographic factors. The first wave of millennials was reaching home-buying age. Population growth was steady. Immigration was contributing to housing demand in many regions.

What happened to supply? Less than demand. Construction did increase — homebuilders built about 2 million new units a year in 2005, an unusually high number — but supply takes time. New housing supply has long lead times (years from purchase to completion), and even at its peak, the supply increase was smaller than the demand increase.

The net effect was that demand shifted right faster than supply shifted right. The model predicts: equilibrium price rises, equilibrium quantity rises. That's what happened — both prices and the quantity of homes sold rose substantially from 2003 to 2006.

The 2006–2012 collapse: demand crashes, supply traps

What happened to demand starting in 2006? Almost everything reversed.

Demand shifter 1 — Lending standards tightened. Starting in 2007, banks (and the regulators behind them) realized that the loose lending of the past few years was producing high default rates. Standards tightened sharply. Borrowers who would have qualified for a mortgage in 2005 could not qualify in 2008.

Demand shifter 2 — Speculative demand collapsed. As soon as prices started to fall (or even just stop rising), speculative buyers exited the market. The expectations that had supported the run-up flipped: instead of expecting prices to rise, buyers expected them to fall. Current demand shifted left.

Demand shifter 3 — The financial crisis. The collapse of Lehman Brothers in September 2008 froze short-term funding markets. Banks stopped lending to almost anyone. Even creditworthy borrowers struggled to get mortgages. Demand for housing essentially evaporated for several months.

Demand shifter 4 — Recession and unemployment. Unemployment rose from about 4.5% in late 2007 to 10% in late 2009. Households that had lost income or were worried about losing income did not buy houses. Demand shifted left.

What happened to supply? This is where the story gets interesting and where the simple supply-and-demand model has to be supplemented.

In the simple model, when demand falls sharply, the price falls and the quantity falls. The market reaches a new, lower equilibrium. End of story.

But housing markets have a peculiar feature: sellers don't always sell at the new equilibrium price. Many homeowners who needed to sell — because they were moving for a job, getting divorced, having children, retiring — found themselves underwater: their mortgage was for more than the house was now worth. Selling at the market price would leave them owing money to the bank with no house to show for it. Many of them simply stopped trying to sell. They held the house, even though it would have been more efficient (in narrow economic terms) to sell.

This is a phenomenon called price stickiness. The supply curve in the housing market is not as responsive to lower prices as the simple model suggests, because home sellers have psychological and contractual reasons not to sell at a loss. Some economists describe this as loss aversion (the behavioral concept from Chapter 10): people hate realizing losses much more than they like equivalent gains, so they hold losing positions longer than the rational model would predict.

The result was that the housing market in 2008–2012 had much lower sales volume than the simple model would predict. People who could have sold at lower prices chose not to. People who needed to buy (especially first-time buyers) found that the few sellers who were willing to sell were demanding prices that didn't match what buyers could afford. The market essentially froze for a few years.

When supply doesn't adjust freely, the equilibrium concept loses some of its predictive power. Foreclosures eventually forced some properties onto the market regardless of seller preferences (the bank doesn't have psychological reasons to hold a bad property), and these foreclosed properties became the supply that did adjust to lower prices. By 2012, after the foreclosure wave had washed through, the market reached a new lower equilibrium, and prices started to recover.

What the model gets right (and what it misses)

What the model gets right:

• The direction of price changes (up in 2003–06, down in 2006–12, up again after) matches what shifts in demand and supply would predict.
• The role of lower interest rates as a demand shifter is well-captured by the model.
• The role of speculation and expectations is captured (as expectations are a shifter).
• The recovery from 2012 forward is well-described as supply-and-demand rebalancing as the financial system healed and household incomes recovered.

What the model misses (or only partially captures):

• Price stickiness in housing supply (sellers refusing to sell at losses)
• The role of credit conditions as separate from price (when credit is unavailable, even buyers willing to pay the equilibrium price can't transact)
• Foreclosures, which create a kind of forced supply that doesn't fit neatly into the simple model
• Distributional effects (the homeowners who lost the most were not the ones who had been most speculative; they were often working-class buyers who had bought reasonable houses with subprime mortgages)
• The interaction with the financial system, where falling housing prices triggered the larger financial crisis

The model is a useful starting framework. To explain the full episode, you also need behavioral economics (Chapter 10), the financial system (Chapter 28), the theory of asset bubbles (which is mostly an advanced topic), and the macroeconomics of recessions (Chapter 30). Each chapter we add gives us more tools for understanding what happened and why.

What the episode tells us about supply and demand

Lesson 1 — Demand shifts can be enormous. The combination of low interest rates, loose lending, and speculation produced one of the largest demand shocks in the history of the U.S. housing market. The model handles this; it just predicts large effects, which is what happened.

Lesson 2 — Supply often takes time to respond. New housing construction takes years from purchase to completion. The supply curve doesn't shift right overnight even when builders desperately want to build more. This is a feature of "long-lived" markets and is one reason housing prices can rise sharply before construction catches up.

Lesson 3 — When supply has frictions (price stickiness, lending constraints), the market can fail to reach equilibrium quickly. This is one of the cases where the simple supply-and-demand model needs to be supplemented with additional concepts.

Lesson 4 — Markets that look like they're in equilibrium can be wrong about future conditions. In 2005, the housing market was in equilibrium — there were no shortages or surpluses, prices reflected what buyers were willing to pay and sellers were willing to sell at. But the equilibrium was based on assumptions about credit, expectations, and lending standards that were unsustainable. When those assumptions broke, the equilibrium broke. The lesson: markets can be in equilibrium and still be heading for trouble.

Lesson 5 — The model is most useful as a first tool, not the only tool. Real episodes — especially dramatic ones — usually involve features the simple model doesn't capture. The full explanation of the 2006–2012 housing collapse requires models from later chapters. But supply and demand is the right place to start, and it correctly identifies the major forces (lower rates → demand right → prices up; tightening credit → demand left → prices down).

Discussion questions

1. The case study identified four reasons demand shifted right in 2003–06 and four reasons demand shifted left in 2006–09. Which of these would you predict to recur in any future housing boom-and-bust cycle? Which were specific to that period?
2. The case study mentioned "price stickiness" — sellers refusing to sell at losses. Why might this be irrational in the strict economic sense but understandable given how human beings actually think?
3. If you were a policymaker in 2008 watching the housing market freeze up, what could you have done to "unfreeze" it? Apply the model to evaluate possible interventions.
4. The case study notes that the 2003–06 run-up was historically extreme but the market still looked like it was in equilibrium at the time. How would you tell whether a current housing market is in a sustainable equilibrium or in a bubble that will eventually burst?
5. The Case-Shiller index is the most-cited measure of national home prices. Look it up on FRED. What does the chart since 2012 show? Is the current housing market reminiscent of 2005 in any ways? Different in others?