Housing Standards 2003/2004:
Housing Policy in the Czech Republic - More Efficiently and More Effective

4. How quickly should rent be deregulated in the Czech Republic (with respect to the efficiency of public housing expenditures)?

4.1 Methodology and assumptions

The rent growth simulation was conducted only with the goal to create annual limits for rent increases. In other words, the analysis says what the maximum increase/decrease of rent in the following one year should be in terms of public expenditure efficiency and therefore the simulation will treat all the other factors as constant (household income, household composition, etc.).

For the simulation of the rent increase and the estimate of costs for the housing allowance, we used the 2002 Family Budget Survey (FBS) containing data on the income and especially expenditures of households. Testing was done on a set of regulated rental housing sector.

When the rent in the regulated rental housing sector increases, a portion of households will move to the owner-occupied housing. This is so especially due to the fact that it will "pay off" to such households to acquire own dwelling despite the fact that the state will provide housing allowances. Three basic conditions had to be met in order for a respondent at a given rent level to move from gradually deregulated rental housing into owner-occupied housing (in the same location, of the same quality and size):

1. it has paid off to the household because the user costs of owner-occupied housing (see below) are lower than the actual simulated rent reduced as a result of eligibility to a potential housing allowance by such a household;
2. the household meets the solvency criteria for receiving a mortgage credit to finance the purchase of its own home;
3. the household is among those that would most likely move from their current home.

The economic aspect of the "tenure choice" theory starts from a decision whether it is more advantageous for a given household to lease or purchase a dwelling. For this purpose, a household compares the net rent costs (i.e., rent after deducting a potential housing allowance) and "user costs" of owner-occupied housing. The standard calculation for user costs (UN) adjusted only according to the specificity of mortgage financing in the Czech Republic, which is usually provided to a maximum of 70% of an estimated price of real estate, was as follows:

where t is a marginal tax rate of a member of a household with the highest income, i is the nominal mortgage credit interest rate, is the rate of wear (funds allocated for the repair fund), is the property tax rate, g the expected rate of price appreciation of given property in the future (i.e. capital growth) and i₀ opportunity costs related to the usage of one's own savings to cover 30% of the property price. In other words, annual user costs consist in the average amount of an interest paid on mortgage credit per year (and potential return on potential use of one's own savings for another investments, i.e., purchase of bonds) reduced by the possibility to deduct from an income tax assessment basis the interest on the mortgage credit (we presumed that the tax deduction will be claimed by that household member receiving the highest income). In addition to the total of the adjusted interest paid on the credit, the user costs also include an estimated contribution to the repair fund (including, also, for example, insurance) and the property tax. Subsequently, the expected price appreciation of property was deducted from this total.

For the purposes of the simulations, i was defined as the average interest rate on mortgage credits provided for housing in 2002 (6.7%) reduced by three percentage point according to the terms offered by the largest mortgage credit provider (the resulting rate depends on the amount of the principal), i₀ is the average interest rate on long-term government bonds in 2002 (4.5%), in an amount of 1%, t according to current income tax rates in 2002 and was excluded from the equation due to its marginal importance. The estimate of the expected price appreciation of property g was performed separately for eight geographical zones and oscillated around 1.5% in Prague and was close to zero in regions with the lowest supply market rent.

The price of a dwelling P^e in 2002 was estimated based on testing hedonic price (OLS) regression models on the data of the Ministry of Finance (MF) and the Czech Statistics Office on purchase prices of dwellings in 2002 acquired from property transfer tax statements.

The first monitored cost item were expenditures for the housing allowance^[1] calculated as a percentage of households eligible to the housing allowance (100% take-up of the housing allowance was presumed) according to FBS 2002 for a given simulated rent level multiplied by the total number of households living in rental dwellings and the average simulated housing allowance for a given rent level. Logically, it was assumed that new households which will acquire vacated rental dwellings will be of the same social and income composition as those that remained in the regulated rental housing after an increase of the rent. According to the presuppositions of the model, the housing allowance was adjusted to cover a full increase of the rent in households that are eligible to the allowance today.

Furthermore, the public costs also included "fictitious" operating subsidies for municipal rental housing amounting to the difference between the cost rent and the actual simulated regulated rent. The amount of the cost rent in existing regulated rental dwellings was defined as 2.8% of the repurchasing value of a dwelling per year. The repurchasing value of a dwelling was defined based on the amount of the current market price of a given dwelling. The operating subsidy was calculated only for municipal regulated rental dwellings and not for private rental dwellings.

Moreover, the costs of public budgets also included subsidies for construction of new rental (e.g., municipal) dwellings. In order to define the amount of these costs, it was necessary to presume a certain normative of neediness of new supported housing construction. For the purposes of the simulation, the neediness normative was defined as the number of dwellings that would satisfy 10% of the estimated unsatisfied demand for housing in a given region. According to the assumption of the simulation model, the demand is formed by households on a common budget living in one dwelling with another household on a common budget or more households and which also stated that the cohabitation with another household posed a problem for them. Also, the additional demand included those households on a common budget sharing a dwelling with one or more households that stated in the survey that they had too small dwelling (or were faced with a lack of space).

In each step (i.e., after each rent increase), the number of households forming additional (unsatisfied) demand first decreased by those households that would acquire dwellings from the pool of the vacated rental dwellings (and the higher the rent level, the greater the number of households that would be satisfied from the existing vacated dwellings) and only then the costs for construction of a number of dwellings equalling 10% of the "remaining additional demand" were calculated.

The model presumed that the rent in new dwellings will be higher in absolute terms (because the reproduction acquisition value of these dwellings is higher) but that it will be defined similarly to other regulated rental dwellings, i.e., that the percentage of the reproduction acquisition value will be the same as in the existing regulated rental dwellings. Even thus "increased" rent, however, does not necessarily cover the costs related to the construction and operation of such dwellings.

The difference between rent paid in new subsidised rental dwellings and the "cost" rent would have to be naturally paid from public budgets, either in a form of irrecoverable capital subsidy (i.e. grant) or a qualified credit. The costs of public budgets following from grants and qualified credits were expressed as their current net value in order to clearly distinguish the costs of the state spent on the grant from those spent on the qualified credit.

For the purpose of determining the lowest costs for the state arising from potential new construction, a financial optimisation programme was developed which sought an optimal combination of a grant, qualified and commercial credit for new construction of municipal dwellings at a given rent level while ensuring the lowest public costs.

In addition to the above-mentioned costs, the model also included so-called rental housing residualisation costs. Those costs include especially residual costs related to the loss of rent in vacant dwellings. In order to calculate the loss of rent in vacant dwellings, we used the simple assumption that today there are no vacant dwellings and that vacant dwellings will appear only when the number of vacated municipal dwellings due to moving out will exceed the estimated additional (unsatisfied) demand in a given region.

The last considered cost to the public budgets in connection with a rent increase were the costs following from pension and social benefit adjustment. According to the calculations, a ten-percent rent increase could result in additional costs following from the pension and social benefit adjustment amounting to CZK 463 million. Although the State would not probably adjust if the consumer prices went up "only" as a result of a ten-percent rent increase (the adjustment would occur only at a moment of a more significant price increase), these costs were taken into account in the calculations.