East European Regional   

34

Housing Sector Assistance Project   

The next step is to design a sampling frame.  At a minimum, the data should be

representative on a nationwide basis.  It may be desirable, however, to also have Poland's

largest cities be represented by a valid statistical sample and perhaps each voivodship.

Similarly, differences between urban and rural households should be distinguished.

The survey should be designed to be ongoing. The AHS is undertaken every other

year, as noted above.  The initial survey will provide a cross section database.  

As time

goes on, this will then emerge into a combined cross section/time series database, which

is very important to understanding a wide variety of dynamic phenomena, such as the rates

and determinants of mobility, life cycle asset behavior and savings behavior, the probability

of becoming a homeowner, the probability of obtaining mortgage credit, and so forth.

#

Develop a Database from Bank Portfolios.

  At present, only  anecdotal 

information exists in Poland regarding mortgage credit aggregates and the details of the

loan portfolios.  The National Bank of Poland does not yet release data that distinguishes

mortgage lending from other types of credit.  The Polish Banks Association, as well, has

not yet made public any information relating to mortgage credit.  During the preparation

of the report  Building on Progress: The Evolution of Housing Finance in Poland, 

information on some of the items note above was obtained (for 1996) through interviews

with the main lenders by UIC team members.  For 1997, estimates of a few aggregates

have been made by CREI staff, with the realization that these figures are basically

 guesstimates.   (These are the figures presented in this report in section 2.0.)

In order to determine whether the information in the portfolios of the lending banks

would support  database creation, CREI staff undertook brief reconnaissance interviews

with the major mortgage lenders.  The results are presented in annex III.  The basic issues

to be addressed were:

!

What data are typically collected by banks during the process of loan

application, underwriting, loan approval, and establishment of the loan in a

portfolio for servicing?

!

Are any or all of these data computerized?

The results suggest that an analysis of existing bank portfolios with regard to their

mortgage lending activity would be an extremely useful first step in understanding the

evolution of mortgage credit in Poland.  This is especially so until a complete household

survey, as suggested above, has been undertaken.  Bank portfolios cannot be used to

obtain (unbiased) estimates of the demand for mortgage credit, since the files contain only

those who have already applied and who have been accepted for a loan.  (In other words,

that portion of Polish households who do not need to apply or do not wish to apply for a

loan or who were refused, which is an interesting analysis in itself are not included.)

Nevertheless, descriptive analysis of current mortgage portfolios would provide very