Integrated System for Demographic Investigations

Transcribed from the printed issue; not yet reviewed. Read it in the PDF of the issue, page 85

The system is intended for specialists, who deal with investigations of demographic processes and use personal computers in their work.

Demographic investigations, and in particular forecasting the growth and movement of population, nowadays play a very important role in a practical solution of economic problems. For example, the knowledge of demographic situation in a region and of the direction of its possible changes is necessary to work out the regional programs of economic and social development. The territories for which these investigations are usually made may differ in size ranging from part of a town to the whole country.

The most important things in demographic investigations are to find out the factors which have an influence on the values of birth and death rates and tendencies of their changes, and to find out the connection between the structure of population and its growth, on the one hand, and migration processes, on the other.

Demographic forecast, especially the forecast of family structure, is necessary for planning housing construction and education development. For example, regional government need to know when, where and how many schools they have to build.

The scientists of the North-West Branch of the Central Research Economic Institute at the State Planning Committee of RSFSR have been carrying out demographic investigations during many years. According to our experience, obtained from practical experience, is that a good demographic forecast is the most accurate among other economic and social forecasts.

The reason for this accuracy is that the methods for demographic calculations have been developing during quite a long time. They are well-known now and permit the scientists to obtain maximum information from poor initial data. The demographic methods are easy to understand and describe the real processes quite well. Also, the demographers are, as a rule, quite experienced in the field of their investigations.

The central planning organs began to use computers for demographic calculations many years ago, but at that time the computers used were big and inconvenient and it was very difficult to use them in scientific research. Now the opportunity to use personal computers gives the new quality to this kind of work.

The first version of the system for demographic investigations which already has been carried out is based upon the well-known method of ages movement. In fact this method is the simplest one and permits to obtain the most accurate results in the forecasting of sex-age structure of population.

The system has three important features:

1. The structure of the input and output data files is made the same as in STATGRAPHICS. Such file organisation allows us to make high level statistical investigations preparing the information for demographic calculations, in particular, making forecasts of birth and death rates changes, people migration movement and so on. It also permits us to present the results of investigations in good graphic form.
2. The system has an exit to APL and return to the system without reloading the workspace. It appeared that when solving a real problem it is more convenient to prepare input matrixes by using APL itself than with the help of a numeric editor, for example, because it’s impossible to formalise in detail the process of preparing information.
3. The system can be easily expanded by including in it new functions which arise during the process of investigation. It was the very reason why APL was chosen as the main programming language for the system.

The system which we are using now in our work consists of the following blocks: the block of preparing input information, the block of calculations, the results scanning block, the printing results block and the graphics block.

The block of preparing input information is used to operate with data files, to create, to copy and to delete files and to edit the information in data files. It also enables us to calculate the coefficients of birth and death rate using the retrospective information.

The calculations block makes it possible to calculate the sex-age structure matrix of population for a given number of years using the ages movement method. It includes the calculation of arbitrary ages groups of population, the average life time and the summary coefficients of birth rate.

The result scanning block enables us to output on the monitor screen the table of the summary population characteristics such as the number of born, the number of dead, the natural and migrational increase, general increase and population size at the beginning of the year, as well as the corresponding average characteristics for the year. It also allows us to scan the sex-age structure matrix on the whole or in fragments and to output on the screen the birth and death rate coefficient matrixes.

The printing block makes it possible to output the same kind of results to the line printer using simultaneously the most convenient printing parameters: font, interval and the number of copies.

Finally, the graphic block permits us to output to the screen and to the printer the demographic pyramid according to the 5-year or random age intervals. All other forms of graphic result representation are supposed to be executed by means of STATGRAPHICS.

This version does not realise everything that we consider necessary even at the moment. Particularly, we have not yet realised the calculations of the population family structure and some service functions. Nevertheless, this version can be considered completed in principal due to two reason. First, it makes it possible to computerise the minimum calculations necessary for demographic investigations. Second, it has already become evident that it is necessary to rewrite some parts of the system if not the most part of it. In this connection a few words might be said about the problems we had to face during our work.

The first problem is that APL is so powerful a language that the programmer is tempted to begin programming the system right after the idea has occurred to him without studying beforehand the system structure and its component parts. That this way is a false one becomes evident when the most part of the system is already completed. Beginning with a certain moment it becomes impossible to introduce any corrections because the system begins to fall apart. Our experience convinces us that even having at one’s disposal such powerful means as APL it’s necessary to accurately study the details at the project stage.

The second problem, which is probably typical for the Soviet Union only, is that the new APL versions are greatly delayed there. As a result you usually find out that the problems that took much of your effort when programming the system could be easily and gracefully solved with the help of the new version. You begin to feel your backwardness acutely and you have an irresistible wish to rewrite everything. If the programmer cannot resist this wish he drives himself into the mode of continuous rewriting the same system version.

Still we hope that, since the problems are clear enough, they are quite solvable.