WaywardHouse
The Bid-Rent Model
Why does a square metre of office space in downtown Toronto cost ten times more than one twenty kilometres away? The bid-rent model gives a precise mathematical answer — and in doing so explains the entire land-use gradient of the modern city.
Prerequisites: Bid rent functions, Spatial equilibrium, Differential calculus
Stand at the corner of Bay Street and King Street in Toronto’s financial core and you are at one of the most expensive parcels of land in Canada. Walk thirty minutes north and the rent per square metre has fallen by more than half. Walk an hour further and you are in low-density residential suburbs where land is cheaper still. Continue into the exurban fringe and you reach farmland. The entire gradient — from glass towers to cornfields — follows from a single economic principle: each land user bids for each location based on what that location is worth to them, and the highest bidder wins.
The bid-rent model formalises this intuition. It was first developed by Johann Heinrich von Thünen in 1826 to explain the concentric ring structure of agricultural land use around a central market town. William Alonso adapted it in 1964 to the modern city, and Richard Muth and Edwin Mills extended it to include housing consumption choices. Together they produced the canonical monocentric city model, which remains the workhorse of urban economics more than half a century later.
1. The Core Idea
Every potential land user — an office firm, a retailer, a household — values central locations more than peripheral ones. The reason is transport cost. Being close to the city centre (or to customers, or to other firms) reduces the cost of moving goods, workers, and information. A firm willing to pay more for centrality can outbid rivals and occupy the central location. Firms less sensitive to transport costs — those that do not depend on face-to-face contact or frequent delivery — are pushed outward by the bidding process.
The bid-rent function $R(d)$ is the maximum rent a given user is willing to pay at distance $d$ from the centre, holding their utility or profit constant. It is not the rent they actually pay — it is their ceiling. The actual pattern of land use is determined by which user category has the highest bid at each location.
2. The Von Thünen Foundation
Von Thünen imagined a featureless plain with a central market town. Farmers must transport their produce to market. The rent they can pay falls linearly with distance because transport costs rise with distance.
For a single crop, the bid-rent function is:
\[R(d) = (p - c) \cdot q - t \cdot d \cdot q\]where $p$ is the market price per unit, $c$ is the production cost per unit, $q$ is output per unit area, $t$ is the transport cost per unit per kilometre, and $d$ is distance from market.
Simplifying:
\[R(d) = q(p - c) - qt \cdot d\]This is a linear function of $d$ with slope $-qt$. Crops with high output per hectare and low transport cost per unit have shallow bid-rent curves. Crops with high transport costs per unit — perishable vegetables, dairy — have steep bid-rent curves and cluster close to the city. Grain, which is cheap to transport, extends far out.
Each crop has its own bid-rent line. At any given distance, the crop offering the highest bid wins the land. The envelope of all bid-rent curves — the upper boundary — determines which crop occupies each distance band, producing von Thünen’s famous concentric rings.
3. Alonso’s Urban Extension
Alonso replaced crops with urban land uses: commercial, residential, and agricultural. The logic is the same but the mechanism differs slightly.
An office firm values central location because proximity to other firms, clients, and workers reduces communication costs. Its bid-rent curve is steep: it gains a great deal from each kilometre of centrality and will pay a premium for it.
A household values proximity to the centre because it reduces commuting costs. Its curve is flatter: households have more locational flexibility and are more sensitive to housing quantity (they want larger lots) than to minute-by-minute access.
Agriculture has the flattest curve of all: farm output per hectare is low relative to urban uses, and farms can sustain high transport costs.
The urban bid-rent hierarchy is therefore:
\[R_{\text{commercial}}(d) > R_{\text{residential}}(d) > R_{\text{agricultural}}(d) \quad \text{at } d = 0\] \[\text{slope}(R_{\text{commercial}}) > \text{slope}(R_{\text{residential}}) > \text{slope}(R_{\text{agricultural}})\]The steeper the slope, the more rapidly willingness-to-pay falls with distance, and the more compact that land use type tends to be.
4. The Muth-Mills Household Model
The household version is more rigorous. A household maximises utility over a composite good $z$ and housing quantity $h$, subject to the budget constraint:
\[Y = z + R(d) \cdot h + T(d)\]where $Y$ is income, $R(d)$ is rent per unit of housing at distance $d$, $h$ is housing consumed, and $T(d)$ is commuting cost (increasing in $d$).
At an interior optimum, the household is indifferent across all locations — otherwise they would move. This spatial equilibrium condition requires that utility is equalised across space. Differentiating the budget constraint with respect to $d$ and applying the envelope theorem:
\[\frac{dR}{dd} \cdot h + \frac{dT}{dd} = 0\] \[\frac{dR}{dd} = -\frac{T'(d)}{h}\]This is the key result: the rent gradient equals the marginal commuting cost divided by housing consumption. Households with longer commutes require lower rents to compensate. The steeper the commuting cost function, the steeper the rent gradient. Households who consume more housing (larger families, higher incomes) live further out because they need the cheaper land to afford their preferred housing quantity.
If commuting cost is linear, $T(d) = t \cdot d$, then:
\[R(d) = R(0) - \frac{t}{h} \cdot d\]The bid-rent function is linear in the simplest case, though in more realistic formulations with elastic housing demand it curves.
5. Canadian Evidence: Toronto, Vancouver, Calgary
The bid-rent gradient is clearly visible in Canadian metropolitan areas, though each city has its own character.
Toronto has the steepest gradient of any Canadian city. Commercial rents in the Bay-King financial district exceed $\$1{,}000$ per square metre annually; class-A office rents in Scarborough or Mississauga are below $\$300$. The gradient steepened after the 2000s as downtown employment concentrated in financial services and technology, increasing the value of central proximity.
Vancouver has a compressed spatial structure imposed by geography — mountains to the north, ocean to the west, the US border to the south. Land scarcity amplifies the bid-rent mechanism: even small differences in centrality are capitalised into large rent differences. Metro Vancouver’s rent gradient is steep and shows relatively little flattening at the fringe because the fringe is geographically constrained.
Calgary offers an instructive contrast. The CBD is smaller relative to the metropolitan area. The oil and gas sector’s preference for low-rise campuses and car-based commuting has produced a more dispersed employment pattern, flattening the residential rent gradient relative to eastern Canadian cities. Still, the downtown core premium is substantial: office rents in Calgary’s core run three to four times those in peripheral employment nodes like the Airport Business Park or Foothills.
6. Visualisations
Bid-Rent Curves by Land Use
The chart shows the envelope logic: commercial use dominates from 0–8 km, residential from 8–15 km, and agricultural beyond 15 km. These crossover points are where land use transitions occur in a monocentric city.
Floor-Space Rent by Distance Band: Canadian City
The steep drop from the 0–2 km core to the 2–5 km ring captures the premium for highest-centrality locations. Beyond 5 km, rents still fall but more gradually — reflecting the flatter section of the residential bid-rent curve and the growing influence of suburban employment nodes.
7. The Density Gradient
Rent gradients translate into density gradients. As rent falls with distance, developers build less intensively — thinner floors, more parking, lower plot ratios. The relationship is well described by the negative exponential:
\[D(d) = D_0 \cdot e^{-\beta d}\]where $D(d)$ is employment or population density at distance $d$, $D_0$ is the central density, and $\beta$ is the density gradient coefficient. A steeper rent gradient produces a larger $\beta$ and more rapid density decline. Canadian cities have seen $\beta$ fall over the post-war period as car ownership flattened the gradient — a phenomenon Colin Clark documented empirically as cities motorised.
8. Assumptions and Limitations
The monocentric model rests on strong assumptions: a single central employment node, homogeneous land, no highways or transit corridors, and perfectly competitive land markets. Real cities violate all of these. Toronto’s employment is distributed across multiple nodes (Yonge-Eglinton, Pearson Airport, Scarborough Town Centre). Vancouver’s Burrard Inlet creates sectoral variation. Calgary’s ring road has generated suburban employment clusters that generate their own local bid-rent peaks.
Nevertheless the monocentric model remains useful as a benchmark. Deviations from the smooth gradient reveal where local access advantages, zoning distortions, or infrastructure investment have altered the competitive outcome.
9. Policy Implications
Because land rent capitalises accessibility, any policy that changes accessibility changes rent. A new subway line increases the accessibility of stations along its route, raising bid-rent curves for commercial and residential users at those stations and pushing out the land-use boundary. Zoning that restricts density suppresses the bid-rent envelope artificially: it prevents users who would bid highly for central land from building to the height that would justify their bid, creating a wedge between potential and actual rent that accrues to existing owners.
Understanding bid-rent also clarifies the distributional stakes of transit investment. A new LRT line that increases accessibility in a previously underserved neighbourhood raises rents there. Existing low-income renters may be displaced as higher-income households outbid them. The bid-rent model does not resolve this tension — it makes it visible.
References
Alonso, William. 1964. Location and Land Use: Toward a General Theory of Land Rent. Cambridge, MA: Harvard University Press. https://doi.org/10.4159/harvard.9780674730854
Clark, Colin. 1951. “Urban Population Densities.” Journal of the Royal Statistical Society, Series A (General), 114 (4): 490–496. https://doi.org/10.2307/2981088
Marshall, Alfred. 1890. Principles of Economics. London: Macmillan. https://www.marxists.org/reference/subject/economics/marshall/
Mills, Edwin S. 1967. “An Aggregative Model of Resource Allocation in a Metropolitan Area.” American Economic Review 57 (2): 197–210. (no public URL available)
Muth, Richard F. 1969. Cities and Housing: The Spatial Pattern of Urban Residential Land Use. Chicago: University of Chicago Press. (no public URL available)
Thünen, Johann Heinrich von. 1826. Der Isolierte Staat in Beziehung auf Landwirtschaft und Nationalökonomie. Hamburg: Perthes. English translation: Von Thünen’s Isolated State, translated by Carla M. Wartenberg (Oxford: Pergamon Press, 1966). (no public URL available for original)