Does Fred Harrison’s 18-year real estate cycle pass the tests of: –
Falsifiability (can it make predictions that could potentially be proven wrong?)
Empirical testability
Logical consistency
Explanatory power
Ability to generate novel predictions (not just post-hoc fitting)
Cumulative progress through testing and refinement
Yes, Fred Harrison’s 18-year land/property cycle approach performs quite strongly — and arguably better than most macroeconomic frameworks — against the standard philosophy-of-science criteria listed. Here’s a breakdown of how it stacks up, based on its structure, historical application, and track record:
Falsifiability
Very strong. The model makes specific, bold, risky predictions: house/land prices peak roughly every 18 years, followed by a major correction or recession. These turning points are dated years (often a decade+) in advance. If the predicted peak does not occur within a reasonably narrow window around the forecast year, or if no significant downturn follows, the theory can be (and would be) falsified for that cycle. Harrison has repeatedly committed to precise dates (e.g., 1990 peak/recession, 2007–2008 peak/crash, 2026 peak followed by downturn), which exposes the model to clear disconfirmation.
Empirical testability
Strong. The theory is grounded in long-run historical series of land values, property prices, and economic downturns (going back 200–300 years in some datasets, especially UK and US). It uses observable, quantifiable data (real land/house price indices, building permits, credit expansion, recession dates). Researchers and critics can — and do — go back to historical records (Homer Hoyt’s Chicago land-value series from the 1930s onward, UK land-price data, etc.) to check whether the ~18-year rhythm holds across multiple cycles. Prospective tests are also possible: we can wait and observe whether 2026 produces the forecasted peak and subsequent slump.
Logical consistency
Good to strong. The core mechanism is coherent within its Georgist framework: Un-taxed or lightly-taxed land rents rise as economies grow → speculative borrowing against rising land collateral increases → credit expansion fuels unsustainable construction and asset inflation → at some point debt service burdens + overbuilding + margin calls trigger a reversal → recession ensues → land prices collapse → debts are liquidated → cycle resets.
This is internally consistent and draws on established concepts (Minsky-style financial instability, rent-seeking, collateral cycles). Critics may disagree with the relative weight given to land vs. other factors (monetary policy, technology shocks, geopolitics), but the causal chain does not rely on contradictions.
Explanatory power
Strong for a particular class of events. The model offers a unified explanation for why many major recessions have been preceded by extreme real-estate/land booms (1830s, 1870s, 1920s, 1970s, 1980s–90s, 2000s). It accounts for: The role of land as collateral in credit cycles.
Why construction and real-estate sectors lead downturns
The link between asset-price collapses and wider banking/financial crises
It does not claim to explain every recession (e.g., purely supply-shock or war-driven ones), but it has strong explanatory fit for the debt-deflationary, asset-bubble-led downturns that have dominated since the 19th century.
Ability to generate novel predictions (not just post-hoc fitting)
Very strong. This is one of its clearest advantages over many competing models. Harrison (building on Homer Hoyt and others) made forward-looking, out-of-sample forecasts well before the events: Predicted the ~1990 peak and early-1990s recession years ahead
In the mid-1990s to early 2000s (notably in his 2005 book Boom Bust), predicted a major peak around 2007–2008 followed by severe downturn
These were not curve-fitted after the fact; they were published prospectively. The current 2026 forecast is another genuine novel prediction being tested in real time.
Cumulative progress through testing and refinement
Moderate to good. The basic 18-year pattern was first systematically documented by Homer Hoyt in the 1930s. Fred Harrison integrated it with Georgist rent theory in the 1980s (The Power in the Land, 1983), made prospective forecasts, and has continued refining timing rules (e.g., distinguishing a ~14-year house-price upswing within the broader 18-year land/business cycle). Other researchers (Fred Foldvary, Mason Gaffney, some Georgist scholars) have applied similar logic independently and arrived at convergent forecasts. There is a small but persistent research tradition that tests and extends the idea rather than abandoning it after failures. That said, it remains outside the mainstream, so it has not benefited from the large-scale econometric scrutiny and incremental model improvement that DSGE or other frameworks receive.
Summary assessment
Harrison’s framework satisfies — and in some respects exceeds — the scientific criteria listed more convincingly than core neoclassical macro models (rational expectations DSGE, representative-agent equilibrium models) when judged specifically on major recession forecasting and on the philosophy-of-science standards of bold, testable, prospective predictions. Its main limitations are: Narrow scope (best at explaining land/credit-driven cycles, less so at short-run fluctuations or non-financial recessions)
Still relatively small body of formal econometric testing compared with mainstream macro
Vulnerability to exogenous shocks that disrupt the cycle (wars, pandemics, major policy shifts)
As of February 2026, the 2026 prediction is actively being tested — many analysts, investors, and commentators are watching land/house price data closely this year to see whether the long-heralded peak materializes and whether a significant downturn follows. If it does, the theory’s scientific credibility will rise further; if not (or if the downturn is minor), it will take a serious hit. Overall, on the criteria named, yes — Harrison’s approach passes those tests more robustly than much of mainstream macroeconomics does for the same purpose.
