

MICROLAND


Microland,  a simple 

macro model 

for open economies




MMC Micromacro Consultants      
Bureau for Economic Research, Training and Advice
C.Jolstraat 50, 2584ET The Hague, the Netherlands      
tel +31 (0)70 3523832  fax 3549589
_____________________________________________________________

Micromacro Consultants in brief


Micromacro Consultants (MMC) is a bureau for economic research, training and advice. It was founded in November 1994 by Dr Marein van Schaaijk. Its aim is to make macro economic knowledge accessible to businesses, interest groups and (local) governments by training, research and special consultancy. 

MMC has started with the expertise of the founder, who has twenty-one years of experience in different scientific and management functions. Besides, some thirty senior economists are willing to give their expertise during short periods varying from some hours up to some weeks a year. The bureau hires junior economists on a flexible basis and besides her own office, also has access to a business centre. The aim is to let the bureau grow in the near future to some five economists on a full time basis.

In the view of MMC there is a market for a small bureau, working between micro and macro, that commercially, quickly, accessibly, understandably and independently of government can do its job, and intensively uses the knowledge of the Planning Bureau and the Statistical Office. 


More specifically MMC is involved in three kinds of activities:

1. The bureau advises internationally on the introduction of the economic modelling approach. 

2. In the Netherlands it provides a post graduate training programme in the effective use of a simple practical macro economic model. 

3.The bureau does quantitative economic research on assignment.


More information of each of these three activities is given in a brochure. For more information please complete the coupon on the last page and send it to us or call us.
Microland, a simple macro model for open economies



A model should not be a fortress in which specialists hide themselves, but it should be a handsome tool for non-specialists to get a better understanding of the way the economy works.













This brochure contents a summary and update of papers by Dr Marein van Schaaijk presented at:

International Conference on Islands and Small States, Malta 23-25 May 1991

Third Conference of Association of Caribbean Economists, Santo Domingo, Dominican Republic, 16-20 July 1991

Symposium on Island Economies: policy models for planning development, Mytilene, Lesbos, Greece, 28-30 November 1991

Mini-Conference on Macro Model Building for Small Economies, Suriname 26-28 July 1992

Symposium on Modelbuilding in the Dutch Caribbean, Paramaribo, July 1993

Enclosure: disk with demo of model Microland

Forthcoming:
MAMIABC, a simple macro economic model in spreadsheet
The Microland Model:
towards a simple general basic macro economic empirical model for open economies 
a model as a simple tool for an all-round economist







Contents                                                         



1. Introduction  5

2. Macroabc, a bookkeeping model in spreadsheet  8

3. Macmics 14

4. Seven steps to a macro model 18

5. work in progress: MAMIABC  23



Appendix A.  From Microland to Malta

Appendix B.  Print Macroabc



1. Introduction to The Microland Model



The Microland model is based on Macmic, a model that was originally built for a small open economy. Microland consists of the simple macro economic part of Macmic, without the special micro block for the export sector. The simple Microland model can be a good starting point for an economic model for other small open economies. It then needs a special microblock, tailored to the most important export products of a country. Because it is so simple and practical, Microland can also be a starting point for a macro model for two other group of countries. Bigger developing countries and countries in transition in Eastern Europe. Neither will need a micro block for the export sector. Developing countries need more attention for the subsistence sector. Countries in transition have no suitable time series available. Because Microland is simple and practical one spares time which can be used for tackling special problems in the modelling of the economy of those countries.

A traditional macro economic model is not suitable for a small country. Neither for theoretical nor for practical reasons. The law of the large numbers does not work in the export sector of a small country. Furthermore, for a small country a team of very specialised modelbuilders is too expensive. (This last point holds also for developing countries and countries in transition in Eastern Europa). However, using a new theoretical approach and a handy simple personal computer, macro model use becomes easy. What the word processor means for a writer, is this macro-micro model for a general economist: one of the tools for analysis. This model can be helpful as one of the tools in making the government budget for the next year, making estimates of the influence of the growth of wages or tourism on the tax revenues, simulating the economic effects of international integration etc..

Writing about 'Microland' we think of one of the islands in the Caribbean, like Aruba, Jamaica, the Netherlands Antilles, St. Lucia, Trinidad, one of the Guyanas, or a West-African country like Gambia, Ivory Coast, Togo, or Iceland, Mauritius, the Seychelles, the Solomon Islands, the Fiji Islands, Western Samoa, Nw. Caledonia,  or Malta (the country of the May 1991 International Conference on Islands and Small States). Besides these very small economies the model could also be of interest to bigger developing countries and countries in transition in Eastern Europe.

So actually Microland is a basis for a general software package for macro model construction, and we showed it by using Malta as a testcase in the paper which we presented in Malta. Here, and in the disk enclosed to this brochure we use the phantasy island Microland. The package 'Microland' in the enclosed disk consists of the national bookkeeping programme Macroabc (to use in conjunction with a spreadsheet), a combined demo of model and a solution programme, batch programmes, and dataset for the phantasy island 'Microland'. (We also have demodisks for Suriname, Curacao, Aruba, Microzemy and Malta, ask Micromacro Consultants).

The country Microland does not exist in reality. It is a phantasy island, which has all the characteristics of a macro economy, but it is small, and 90% of the export value is furnished by only one product. Microland has 250,000 inhabitants. The dataset of Microland has been made out of a mixture of the situation in Suriname in the seventies, and some island economies in the seventies and eighties. The parameters of the model Microland have been inspired by Suriname in the seventies. So we built Microland as an intermediate station to different small countries.

In small ("micro") open economies 90%  of the export value is provided by some ten or fewer products. The law of large numbers does not work then, so the usual macro models fail.
  A solution to this problem is the incorporation of a micro block for the export sector. This consists of special price, production and investment equations for every of the main export products. This block has to be tailor-made for every small country, making use of the knowledge of business economists. This micro method would be very expensive for big countries, but is cheap for small countries.
  Not the whole model for a small country needs to be tailor-made. The macro block of the model can be ready-made, because it is for different countries approximately the same. One only needs to reestimate the cofficints, and to construct a dataset, for which our bookkeeping model Macroabc will be helpful. In this paper we will only discuss the macro block. A micro block for the export sector will need big differences for different countries. In this paper we concentrate on the macro block which for different countries basically could be the same.


Our model Microland could be of interest for small ("micro") countries, but only for countries with their own macro economy. It is not the size of a country which defines if it has its own macro economy, or if it is a part of a bigger entity. We define here a macro economy as an economy which has its own government with an independent budget policy, its own monetary unit, its own wage formation process, economic boundaries by physical frontiers and/or import taxes. From this point of view provinces of a country and even independent nations like Luxembourg and Monaco do not have a macro economy, but for example the Netherlands Antilles do have their own macro economy. A small country in our opinion is a country with 50,000 to 2,000,000 inhabitants.

Our macro model theoretically has been founded on the modelling of the export sector on the most desaggregated level, the micro level, actually for each of the products which together contain 90% of total export value. 

The method of modelling the exposed sector on product level is in practice to realise, if the number of export products is limited, while the non-export part of the exposed sector is rather small. This situation may be expected for small, open economies.

The theoretical method of the micro block could be used for different small open economies, but of course the practical composition has to be different for every country, because the essence of the method lies in the special treatment of every export product. Because different countries have different kinds of export products, it is not possible simply to copy the micro block for other countries. For the method of the micro block we refer to 'A Macro Model for a Small Economy' (see last page of this brochure). Here we will concentrate on the application of the macro-block of the model Macmics and the bookkeeping model Macroabc to other countries.

Practically our model has been based on a construction method, starting with information about only 70 primary variables, one can complete a consistent dataset which is needed for an empirical macro economic model. 

In the next section we will present Macroabc, then we will present Macmics and then make the step from Microland to another country in general. 
2. Macroabc, a bookkeeping model in spreadsheet


The most time-consuming part of the construction of the first empirical economic model for most countries forms undoubtedly the construction of the data set. One is confronted with breaks in time series, obscurities and unreported changes in definitions, inconsistencies, gaps, mistakes. Moreover, one has the risk to drown in an abundance of data which appear at the end not necessary, while the data one really needs are missing. An important step forward can be taken if one knows exactly what data one really needs, leaving all other unordered data untouched.
  We are not talking here about the micro data. The relevant micro data differ from country to country. So it is no use talking about them in this generalisation of our model for other countries, apart from a statement which can give some optimism: if a product is really important for a country, then one usually has made statistics about that product already for purposes other than constructing a model!

The primary data file contains only the strictly necessary data. No more no less, for making the complete macro survey of the economy, which is needed as an empirical foundation for a macro economic model.

For the macro data one must realise that in the National Accounts most variables are printed twice, while between many of the variables, definition relations exist. Many variables are deduced from each other. The enormous number of leaves rise from a small number of branches. We have been looking to find the common branches behind the collection of definitions. Doing that one can make different choices. In the same way, the consumption of households can be measured by the households themselves, but also by the delivering enterprises.  We could take advantage of our practical statistical experience in making National Accounts, and construct a primary data set with a selection of only 70 variables, which nevertheless gives exactly enough basis information for Macroabc to deliver all the many hundreds of input data for model and estimation of the behavioural relations. In principle one can make, starting with only 70 variables, something like 70 factorial other variables. That is nearly infinite. The bookkeeping model Macroabc contains 850 rows, and further enlargement is possible. However, even in that case, it is enough if one collects information about only the seventy which one finds in following tables. 
Variables in the primary inputfile for Macroabc:   

		PRIMARY INPUT SECTOR ENTERPRISES:	
			
	1	gross value added enterprises (m)	
	2	gross investment by enterprises	
	3	wages enterprises	
	4	depreciation by enterprises	
			
		PRIMARY INPUT SECTOR REST OF THE WORLD:	
			
	5	export goods+services by enterprises	
	6	import goods+services by enterprises	
	7	import goods+services by government	
	8	import goods+services by households	
	9	wages from abroad	
	10	property and entrepreneurial income from abroad	
	11	wages to abroad	
	12	property and entrepreneurial income to abroad	
	13	unreq. current transf.from abroad to households	
	14	unreq. current transf.to abroad by households	
	15	increase in official reserves	
			
		PRIMARY INPUT SECTOR GOVERNMENT:	
			
	16	goods+services by government to enterprises	
	17	goods+services by government to households	
	18	gross investment by government	
	19	property and entrepreneurial income from enterprises	
	20	indirect taxes	
	21	direct taxes by enterprises	
	22	direct taxes by households	
	23	unrequited current transfers by households	
	24	unrequited current transfers by rest of world	
	25	unrequited capital transfers by households	
	26	unrequited capital transfers by rest of world	
	27	goods+services by enterprises to government	
	28	depreciation	
	29	wages	
	30	interest by government to enterprises	
	31	interest by government to rest of the world	
	32	subsidies	
	33	unrequited current transfers to households	
 33	unrequited current transfers to rest of the world	
			
		PRIMARY INPUT FOR MONETARY SURVEY:	
		  sector rest of the world:	
	35	surplus on current transactions (cash basis)
	36	surplus on capital transactions	
	37	valuation differences etc.	
			
		  sector government:	
	38	surplus on current account (cash basis)	
	39	net development loans	
	40	net private loans	
	41	monetary financing by government	
		 	
		  change in liquidity:	
	42	primary and secondary liquidities	
		 	
   PRIMARY INPUT: PRICES	
	43	consumption price change	
	44	export price change	
	45	import price change	
	46	gross investment price change	
	47	change in government wages	
			
		PRIMARY EXTRA MACRO INPUT FOR MACRO-MODEL:	
	48	jobs in enterprises        
	49	jobs in government        
	50	unemployment rate         	
 51	index trend in labour productivity	
	52	population                
	53	population 16/64 year     
	54	natural growth population 16/64 year	
	55	emigration net            
	56	idem cumulated
	57	interest rate USA	
	58	interest rate in the country	
	59	foreign exchange holdings	
	60	money supply	
	61	public internal debt	
	62	public external debt	
	63	invested wealth of enterprises(rough estimate)	
	64	capital import enterprises cumulated	
	65	export main products 	
	66	gross investment in export sector	
	67	indirect taxes on main exports	
	68	exchange rate  
	69	value of import taxes	
	70	capital from enterprises to government	
Thanks to the bookkeeping model Macroabc, one can concentrate on only seventy variables. Because of the selection to this small number, combined with the possibilities of modern personal computer technology, it becomes possible with very limited manpower to build a complete data set.






How to start Macroabc
 
You need a MSDOS personal  computer, 640Kb, and a spreadsheet program (Lotus, or VPP, or ASEASYAS, or MSWORKS etc.).
 
Start your computer and start your own spreadsheet (for example  VPP). Then we load Macroabc.wks in the  spreadsheet. Type /FR  De  pc   now asks   for  the   name  of   the  spreadsheet.  Type A:  Macroabc<enter>. After  some time  you see  Macroabc coming on the screen. Macroabc contains already all formulas, but yet no input. We read the inputfile: first check if the cursor is in cell A1. If not bring it there. Type then: /F C C E    A: BAS1-9<enter>
Now the input has  been read and  stands on rows 1  to 108. If you push F9 now,  then  are  calculated  together:  Sector Accounts, National  Accounts, constant  prices,  Monetary  Surveys, Labour Market Balances,  etc..  If you  want  to  print  the  results of Macroabc for year 1, type  / P P R a1..b850<enter>,  and  after  the  printer  has  been adjusted on 'condensed' type   G. If you want to leave the spreadsheet type / Q Y Q.  See also appendix B for a print of Macroabc.



Up till now Macroabc has been used as a bookkeeping model to analyse the past. If one should have assumptions or forecasts for the seventy primary variables, then one can use Macroabc to make - by recalculation- the many hundreds macro variables of the National Accounts, Monetary Survey and Labour Market Survey for the future.

So the next question is, how one can make assumptions or forecasts for the seventy primary variables. The first step is to make a division between exogenous and endogenous variables. The exogenous variables are bases on assumptions outside the model. Exogenous are the following nineteen variables listed on the preceeding pages: 9, 11, 13, 14, 23, 24, 26, 32, 33, 34, 37, 39, 40, 45, 49, 51, 54, 57, 70. Endogenous are the other fifty-one of the primary variables. A model whichs makes forecasts for the endogenous variables is Macmics. See the next paragraph.


3. Macmics


Macmics is the name of the macro model of the economy of Microland. The model is written in the solution programme SIMPC and available in the disk to this brochure. Macmics makes forecasts of the endogenous variables, starting with their values in the past and the definition relations like in Macroabc.

Macmics consists of two kinds of equations: behavioural relationships and definition relationschips. We mention the most important behavioural equations. 

Behavioural equations:

Instead of micro block:

The next three macro equations are only temporary used, waiting for the construction of a micro block for the export sector:

The exportprice is a function of import prices and wage costs

The export value is a function of the difference between export and import prices

The investments in industries are a function of gross value added of industries

Macro block:

The change in value of consumption is a function of disposable income of wage earners and profits

The change in quantity of imports is a function of final output and import shares and difference in internal and external inflation

The growth of employment in industries is a function of growth of final output and corresponding labourintensities and the change in unemployment rate

The change in unemployment rate is a function of demand for labour and the increase of labour supply

The change in wage follows the consumption price and unemployment change

The investment price change is a function of the cost development 

The change in consumption prices is a function of cost development and change in indirect tax rate 

Money growth is a function of monetary financing bij government and surplus on the balance of payments 


Some semi-behavioural equations, only about taxes:
(no % changes, but levels)

Importduties is a function of the value of imports 

Taxes on consumption is a function of value of consumption 

Direct taxes on industries is a function of disposable profit income:

Direct taxes on persons is a function of disposable income of employees and profits 




Besides the behavioural equations the model Macmics consists of definition equations. The best way to get an impression of the model is to start the demo in the disk to this brochure.



HOW TO START A DEMONSTRATION WITH Macmics
  
this demonstration will last for only 
one quarter of an hour

Start your personal computer with MSDOS. Then disk in drive A or B. Type then:  macmics<enter>
Wait a  moment and type after "enter command?"  

 ai demoml <enter>

Then the demonstration begins. This demonstration is regularly interrupted, and starts again if you push the <enter> button.

In the demo is shown:

-entering the data file of Microland
-a historical simulation of the economy of Microland
-a baseline, a reference forecast to the future
-policy simulations, for example:  what happens in Microland after an
 impulse of five percent additional increase of the wages, or what
 happens if exportprices have ten percent extra growth
-print of results in tables and graphs

During the demos comments are given. After running the demos you can run your own simulations to get familiar with the model.

If you like, you may start the demos you have seen in the batch mode now in the direct mode. Start the computer and after macmics <enter> you kan type the following commands (read the file h1-9.dmi which contains the data; set the sample period on year 3 until 6; simulate; show a table with key figures; read a base line and compare the results of the simulation with those of the baseline in two tables): 

r h1-9.dmi
smpl 3 6
sim p;y
table kern.tbl
base h1-9.dmi
table kerndif.tbl
table kernpdif.tbl

You have finished the running of a new base line and have looked at the results. If you want to stop, type qq, followed by <enter>, or run for yourself simulations in the direct mode, for example a wage simulation (show lbabppau, an impuls in wage increase; add 5 %; show the result; simulate; read a base line; compare the results of the simulation with those of the base line; set the impuls in the wage increase back to zero):

pr lbabppau
calc lbabppau=lbabppau+5
pr lbabppau
sim p; y
base h1-9.dmi
table kernpdif.tbl
draw dif(rvblp) dif(wwqn)
y
load LBABPPAU; 9*0;

Or run a simulation with 10 % additional increase in exportprices (bppt), and look on the effects on purchasing power (rvblp), increase of employment (abap) and unemployment rate (wwqn):

pr bppt
smpl 3 4
add bppt; 2*10;
pr bppt
sim p; y
base h1-9.dmi
table kernpdif.tbl
draw dif(rvblp) dif(wwqn); y
draw abap lbabpp; y
qq 


Of course you can make all kinds of variations on these demos. That is the way to teach yourself how to work with a macro model. The disk contains besides Microland (or Miniland) als demo models for Suriname, Malta and Curacao (Netherlands Antilles). You can run these demos by -after entering macmics and waiting for a moment- typing Suriname, Malta or Curacao. That is another way of getting familiar with macro models.
However, you can also join one of the five-day Micromacro Training Programmes. You can get information if you send the coupon on the last page to Micromacro Consultants.

The model Macmics can be used for four purposes:

-Analysis of the past, both for scientific purposes, as for the improvement of the parameters of the behavioural relations.

-Short, mid term and long term forecasting. With the model one can make a prognosis if one has information about the values of the exogenous variables, such as some government variables. Often a model is in use in the budget preparation process and then prognosis and filling up the value of the government exogenous often forms an iterative process.

-Simulations. A model is then used as an instrument for civil servants and
politicians to become aware of the consequences of policy intentions and an assistance for policy coordination between Ministries.

-Updating and Monotoring. In practice, the present and recent past form from a statistical point of view something which has yet to be realised. Even the most recent statistics have a delay. One can use a model to construct the provisional National Accounts, starting with the realised Accounts about some years ago, and as many as possible realised data, while the model generates consistently the rest of the variables. 


In this new training programme we use  a new version of the model, MAMIABC, which also runs in a spreadsheet and experience learns that this gives an even easier way of getting familiar with macro modelling.



4. Seven steps to a macro model


In the accompanying diagram, the headlines of the activities which are recommended to construct efficiently a macro model for a small open economy with limited data have been outlined. For the time being, the micro block is left exogenous. The next seven steps can be distinguished:

1. The construction of a primary data set. Using our bookkeeping spreadsheet Macroabc, an efficient data set of only 70 variables is enough to fill a secondary data set of many hundreds of variables, which are needed as input for the macro model Macmics. The selection of the 70 primary variables is so important that we focus on this step in a special paragraph.

2. The primary data set of step 1 gives the input for the book keeping-program Macroabc. This program is of importance for the analysis of the data and the testing of them on consistency, which is also of interest for the construction of the primary data set. So there will be some iteration between step 1 and 2, before making the definitive values of the 70 variables of step 1. Because Macroabc can be used in a spreadsheet program, it can easily be enlarged and/or adapted to the specific needs of the country involved. For example one can easily make changes in Macroabc to adjust it to the definitions in use by the Statistical Office of the country by making the National Accounts, the definitions of the Central Bank for variables in the Monetary Surveys, etc..

3. After the construction of the primary data set and the statistical analysis of it, the search for empirical relations, based on economic theory and knowledge of the institutions of the country, follows. Graphic analysis can be very clarifying, and easy to perform using the facilities of a spreadsheet. Furthermore one can make regression calculations.

4. The analyses in the next step will result in empirical behavioural and institutional parameters for the (semi)-behavioural equations, which together with the definition equations and the data form the model.

5. This step starts with exactly the same 70 primary variables as Macroabc.
Via definition equations all the starting values for the model (excluding
micro-block) are supplied.

MICROLAND DIAGRAM   
1.       primary data file with 70 variables:

34          NationalAccounts
8          Monetary Surveys 5         Prices
23         Labour Market
 2.   MACROABC with: 
    Sector Accounts   
    National Accounts
    Monetary Surveys
    Labour Market etc.
5. translation
to input file for Macmics
3. graphic analyses and regression calculations
6. exogenous variables
   for prognosis years
4. behavioural and institutional parameters


  7. THE MODEL MACMIC



UPDATING,  FORECASTING,  SIMULATION,  ANALYSIS
6. The exogenous variables concern mainly the international  price developments, the demographic, and a number of government variables. In the version  excluding the micro block, the export variables are also exogenous.

7. Data plus model need a solution program to let the model run. A special freely distributable version of a solution program combined with the model, called Macmics, is in the disk.


These seven steps provide a first version of an economic model. With this first version one can make historical simulations and compare these with historical data. In the second place one can also make partial simulations. Furthermore, in the third place, one can make simulations for recent years and compare some of the results with the statistical information  which has recently become available. In the fourth place one can add a micro block for the export sector. These four kinds of model use give plausibility checks. They can give reason to review some specifications and change them. Then you will have an operational model which can be uesed for analysis, forecasting, simulations, updating and monitoring. 

Then you can enlarge the model with satellite models for purchasing power, for government budget, monthly monitoring, energy, etc..

Package 'Microland'

In the disk one finds not only the programmes Macroabc and Macmics, but also a primary input file for the phantasy country "Microland", and a file of that country which is ready as starting data set for simulations. Furthermore the disk also contains demo models for Suriname, Malta and Curacao (Netherlands Antilles).




Possibilities for adapting Microland to another country:


1. Create a primary input file for Macroabc.
2. Run Macroabc with this new primary input file.
3. Analyse the data.
4. Make regression analysis to find values for the parameters
5. Use the new data set as input for model ( or changes of values 
   in the h1-9.dmi file).
6. Make assumptions about the exogenous variables.
7. Make some behavioural equations exogenous in Macmics. Change
   tax parameters via other values for ....AU. Moreover adapt VHMVNT
   and VHAVNT. Fill up for every behavioural equation the relevant CA
   variable wit the residuals from first tests.

Ask tips from Micromacro Consultants and help for a further improved version of your model.  See for an example Appendix A: From Microland to Malta.


Further steps

The package Microland gives the opportunity to do some exercises to get an idea of how to construct a micro model for a small open economy, based on a data set of the country. However, a real model asks for more. You must realise that model Macmic Simple (MacmicS) in the package Microland is only a simplified, demoversion of the model Macmic. Macmic consists of 246 equations, so it contains 246 endogenous variables, of which 57 estimated behavioural variables. Although we tried to make the model as condensed as possible, it appeared not to be possible to make an even simpler and still suitable model. MacmicS contains approximately only one half of Macmic. The exchange rate equation, the micro block etc. are excluded. If one wants to use Microland professionally, one will need a more enlarged version of MacmicS. 
5. Work in progress: Mamiabc


Last years computer technology made a lot of progress and nowadays it is feasable to use a spreadsheet to solve simultanous macro models. Because most economists are already familiair with spreadsheets, a macro model which can run in a spreadsheet is very accessible in practice. That is why we developed a spreadsheet version of Macmics.

This ABC of a MAcro model for a MIcro economy is called MAMIABC. For Suriname we have tested versions of this model and we used it for a training programme for economists of the Ministry of Trade and Industry. For Curacao (Netherlands Antilles) and Aruba testversions are working and we are also working on such a spreadsheet model for the Netherlands.

This Mamiabc has a double purpose. Practical economists who do not work full time with models can use Mamiabc as a tool to work quickly and efficiently with a model. One does not need to learn working with a solution programme. Only for economists who work dayly with a model the investment in time to learn the solution programme is profitable. 
The other purpose of Mamiabc is to give economists in small economies or organisations which do not have not so much time, and specialised economists for economic modelling activities a tool to develop in less time a model for their economy.  Of course Micromacro Consultants is wiling to help them to do this job. This approch saves time, which can be used for completing the model with a country specific micro block for the export sector and satellite models for purchasing power, government budget, monthly monitoring of prices, etc..

The following tables give an impression of Mamiabc. The first diagram shows Mamiabc as a bookkeeping model. That version af Mamiabc is identical to Macroabc, which we discussed before: With 70 primary variables on row 1 to 108 the model makes all other variables via definition equations. In diagram 1 the primary input is indicatd with '123' and the variables accounted on basis of the primary input are indicated with 'abc'. So with the input of 70 primary variables Mamiabc fills variables in 860 rows.

If one can forecast the 70 primary variables, then one has also the other available thanks to Mamiabc. By forecasting the primary variables one can make a distinction between exogenous variables, which one has to estimate outside the model, and endogeneous variables which one can calculate in the model on basis of behavioural relationships. In diagram 2 one sees in the column concerning the future for the first hundred rows (containing the primary variables) that some of them are exogenous (indicated with 123) and some endogenous (indicated with abc). The endogenous variables are calculated from other primary variables and definition relations. In row 861 up to 1039 additional definition variables are made to make calculations more easy. At the end one sees some key figures to give in one small table a view to the economic development.
 
For more information about Mamiabc, the way to make it for a country and how to train economists in using it we refer to a forthcoming brochure of Micromacro Consultants and the folder with the Micromacro International Programme on Practical Macro Economic Model Building, see coupon on the last page of this brochure.



	DIAGRAM 1.   MAMIABC.WKS   as bookkeeping model						
		LEGENDA: 123=amount, abc=formula					
							
		NAME                           PAST					
		                               1993		
	1	_______________________	
		PRIMARY VARIABLES:	
	14	enterprises	|123  |
	u/a		|123  |
	17		|123  |
	21	rest of the world	|123  |
	u/a		|123  |
	31		|123  |
	35	government	|123  |
	u/a		|123  |
	46		|123  |
	60	for monetary survey	|123  |
	u/a		|123  |
	73		|123  |
	79	prices	|123  |
	u/a		|123  |
	83		|123  |
	86	labourmarket etc. 	|123  |
	u/a		|123  |
	108	_______________________	|_____|
		SECTORACCOUNTS	| abc |
	540	_______________________	|_____|
		MONETARY SURVEY	| abc |
	648	_______________________	|_____|
		RESOURCES & EXPENDITURES	| abc |
	702	_______________________	|_____|
		BALANCE OF PAYMENTS	| abc |
	756	_______________________	|_____|
		GOVERNMENT FINANCE	| abc |
	810	_______________________	|_____|
		LABOURMARKET SURVEY	| abc |
	830	_______________________	|_____|
		BALANCE VARIABLES	| abc |
	861	_______________________	|_____|






DIAGRAM 2. MAMIABC.WKS as bookkeeping and prognosis model							
		LEGENDA: 123=amount, abc=formula							
                                      PAST | PRESENT&FUTURE
                                 			
                                     |1994 |1995 |1996 |
	1	_______________________	|     |     |     |
		PRIMARY VARIABLES:	|     |     |     |
	14	enterprises	|123  | abc | abc |
	u/a		|123  | abc | abc |
	17		|123  | abc | abc |
	21	rest of the world	|123  | abc | abc |
	u/a		|123  | abc | abc |
	31		|123  | abc | abc |
	35	government	|123  | abc | abc |
	u/a		|123  | abc | abc |
	46		|123  | abc | abc |
	60	for monetary survey	|123  | abc | abc |
	u/a		|123  | abc | abc |
	73		|123  | abc | abc |
	79	prices	|123  | abc | abc |
	u/a		|123  | abc | abc |
	83		|123  | abc | abc |
	86	labourmarket etc. 	|123  | abc | abc |
	u/a		|123  | abc | abc |
	108	_______________________	|_____|_____|_____|
		SECTORACCOUNTS	| abc | abc | abc |
	540	_______________________	|_____|_____|_____|
		MONETARY SURVEY	| abc | abc | abc |
	648	_______________________	|_____|_____|_____|
		RESOURCES & EXPENDITURES	| abc | abc | abc |
	702	_______________________	|_____|_____|_____|
		BALANCE OF PAYMENTS	| abc | abc | abc |
	756	_______________________	|_____|_____|_____|
		GOVERNMENT FINANCE	| abc | abc | abc |
	810	_______________________	|_____|_____|_____|
		LABOURMARKET SURVEY	| abc | abc | abc |
	830	_______________________	|_____|_____|_____|
		BALANCE VARIABLES	| abc | abc | abc |
	861	_______________________	|_____|_____|_____|
		HELPVARIABLES	| abc | abc | abc |
	890	_______________________	|_____|_____|_____|
		DEFINITIONS	| abc | abc | abc |
	1039	_______________________	|_____|_____|_____|
		KEYFIGURES	| abc | abc | abc |
	1059	_______________________	|_____|_____|_____|
			




Appendix A    From MICROLAND to MALTA



The model Macmics for Microland in the disk to this brochure can be transformed to a model for another country. We show this for Malta. The only thing you need is a data file for some years of your country. Then you use Macroabc to analyse the data and then you reestimate the behavioural equations. After that you change the parameters in Macmics, and there you are: you have a testversion of a model of your country. Let us see how we did this for Malta:



Statistical sources for Malta:

For Malta we used the following statistical publications:
-Central Office of Statistics-Malta:"National Accounts of the Maltese Islands 1989" Idem 1984 and idem 1968-1977
-Central Office of Statistics-Malta:"Abstract 1989"
-Central Bank of Malta:"Quarterly Review December 1990"
-Economic Planning Division, Ministry for Economic Affairs Malta:"Economic Survey January-September 1990  (The National Accounts Division of the Statistical Office gave us also a copy of the "Main economic indicators" for the years 1966-1985).
-Lino Brugiglio:"The Maltese Economy", Malta,1988.

Thanks to the well organised statistics, especially the National Accounts, of Malta we could collect the data in a visit of less than three hours to Valetta. Four evenings working were enough to fill up the primary data file with its seventy variables. Ten seconds later we had the whole Macroabc for Malta for the years 1972-1989 available. 

Macroabc deliveres all the data which are needed to estimate the behavioural equations.


Main behavioural equations estimated for MALTA
(excl. micro block):   


%changes                        estimation period: 1974-1989


Instead of micro block:

The next three macro equations are only temporarily used, waiting for the construction of a micro block for the export sector (T-values in parenthesis):


The exportprice(BPP) is a function of importprices (MPP) and wage costs (HPEPWP):

BPP=  .9  +.6*MPP  +.2*HPEPWP                    RR=.92 DW=2.1
      (1.4) (10.6)   (1.8)

The export value as a function of the difference between export and import prices:

BWP=  10.9  -2.4*(BPP - MPP)                     RR=.37 DW=.9
      (3.6) (2.8)

The investments(IBWP) in industries as a function of gross value added of industries:

IBWP= 4.8   +0.8*YBWP                            RR=.15 DW=2.1
       (.6)  (1.6)

It appears not to be possible to make a good investment equation without a micro block for the exportsector, with a breakdown of exports to tourism, ship building etc.

Macro block:

The change in value of consumption(CGWP) as a function of disposable income of wage earners (LBESQW) and profits (ZBESQW):

CGWP=  2.7   +69*LBESQW +16*ZBESQW               RR=.48 DW=2.4
      (1.0)  (2.6)      (0.6)
The change in quantity of imports as a function of final output and importshares(VHMVP) and difference in internal and external inflation (CGMPP=CGPP-MPP):

MVP= -2.9    +1.3*VHMVP +.2*CGMPP                RR=.64 DW=1.3
      (1.3)  (4.8)      (.8)


The growth of employment in industries(ABAP) as a function of growth of final output and corresponding labourintensities (VA), and the change in unemployment rate:

ABAP= -0.9 +0.59*VA(-1) -0.23*WWQD               RR=.28 DW=1.5
      (.4)  (1.9)        (.2)       (simultaneously with WWQD)

The change in unemployment rate (WWQD) as a function of demand(ABAP) and demand (POTBAP):

WWQD= -.3  -.23*ABAP +.64*POTBAP                 RR=.58 DW=2.5
      (.5)  (2.1)     (3.9)         (simultaneously with ABAP)

The change in wage(LBABPP) followes the consumptionprices (CGPP) and unemployment change:

LBABPP= 2.8  +1.3*CGPP(-1) -1.3*WWQD             RR=.55 DW=2.2
       (1.3)  (4.0)         (1.3)

The investmentprice change (IPP) as a function of the cost development (HPMI=.54*MPP+.46HPEPWP):

IPP=    1.8  +.9*HPMI                            RR=.54 DW=1.7
        (1.0)  (4.0)

The change in consumption prices as a function of cost development (HPMC=.27MPP+.73HPEPWP) and change in indirect tax rate (TKQ)

CGPP=   1.3  +.55*(HPMC TKQ- TKQ(-1))            RR=.60 DW=2.3
        (1.2)  (4.6)

Moneygrowth as a function of monetary financing bij government (FINWD) and surplus on the balance of payments (RESWD):

LIQWP=  6.3  +.6*FINWD +.6*RESWD                 RR=.62 DW=2.1
        (5.5)  (1.7)    (4.6)

Some semi-behavioural equations, only about taxes:
(no % changes, but levels)

Importduties (TMWN) as a function of the value of imports (MWN):

TMWN= -2.0 + 0.08*MWN                            RR=.96 DW=.8
       (1.2)  (17.7)

Taxes on consumption(TCWN) as a function of value of consumption (CGWN):

TCWN= -.6  + 0.07*CGWN                           RR=.97 DW=.7
        (.7)  (21.0)

Direct taxes on industries(TDBWN) as a function of disposable profit income:

TDBWN= 1.2 + 0.06*(ZBESBWN(-1)+TDBWN(-1) )       RR=.93 DW=.7 
       (1.5)  (13.7)

Direct taxes on persons (TDGWN) as a function of disposable income of employees (LBESBWN) and profits (ZBESBWN):

TDGWN= 2.9 + 0.11*(ZBESBWN(-1)+LBESBWN(-1))      RR=.93 DW=.50
       (.8)  (13.2)


NAMES OF VARIABLES IN MAIN BEHAVIOURAL EQUATIONS:

ABAP     number of jobs in enterprises % growth
BPP      export price %
BBPMWP   value gross domestic product %
CGPP     consumption price %
CGWN     value consumption households
CGWP     value consumption households %
CGMPP    consumption price % minus import price %
FINWD    financial shortage of government in % moneyvolume t-1
HPMC     function of wage costs and import costs
HPMI     function of wage costs and import costs
IBWP     value investment enterprises %
IBMWPQ   value investments in micro block % in share total
IPP      investment price %
LBABPP   growth of wages per caput in enterprises
LBESBWN  disposable wage income
LBESQW   disposable wage income % share in total
LIQWP    money growth %
MVP      volume imports %
MWN      value of imports
POTBAP   growth of potential labour force excluding civil servants
RESWD    surpluss on balance of payments in % money volume t-1
TCWN     taxes on consumption
TDBWN    direct taxes on enterprises
TDGWN    direct taxes on households
TKQ      tax pressure
TMWN     import duties
VA       final output reweighed by labour intensity minus
         number of jobs, divided by number of jobs year before
VHMVP    final output reweighed to import intensities
WWQD     unemployment rate this year minus year before
YBVP     volume gross value added enterprises
YIWP2    value gross value added enterprises excluding
          exports, half a year delay
ZBESBWN  disposable income profits
ZBESQW   disposable profit income % in share total

....AU   autonomous or correction factor


Historical simulations for Malta.


This data set made for Malta using Macroabc has been used as input in the model Macmics, instead of the data set of Microland. Furthermore, a historical simulation has been made for the period 1975-1980, starting with the data of the year 1975 and for the years 1975-80 using only information about the exogenous variables (mainly international prices, natural growth of the population, value of government consumption). This is also done starting in 1983 for the period 1983-1989. Subsequently, the constant factors in the behavioural equations have been enlarged with the average difference between the result of historical simulation and realisation figure of the relevant dependent variable. Furthermore, the parameters of the tax equations have been improved. After this 'calibration' we made again simulation from year to year, starting with the data of the year 1975, and for later years up to 1980 only information about the exogenous variables.We made such a historical simulation also for the period 1983-1989. These simulations appeared to be technically successful.  Of course this technical success does not give a model for Malta. Practical usefulness requires more than this 'calibration'. The macro block needs further development in particular the export equations need to be improved. Because of these and other shortcomings we must stress that this test version is to be considered as a preliminary one.

See for more information about this subject the brochure 'A macro model for Malta', see coupon on last page.

MMC  Micromacro Consultants      
Bureau for Economic Research, Training and Advice
C.Jolstraat 50, 2584ET The Hague, the Netherlands      
tel +31 (0)70 3523832  fax 3549589
_______________________________________________________


COUPON FOR FREE FOLDERS AND FREE DEMO DISK

Complete this coupon and send it to the address above

Name.....................................

Institution..............................

Address..................................


Yes I want info about Micromacro Consultants and its 
O  Micromacro Training Programme
O  International Programme on Practical Macro Economic
    Modelbuilding
O  Research on Assignment 

O  Yes I want the free demo disk with the model 'Microland'

O  Yes I want information about new activities and brochures of MMC

You may send such a brochure about Micromacro Consultants also to my colleague/friend/relation:

name..........................

address........................

Please send me a copy of the next brochure:

O A macro model for a small economy, 
  ISBN 90737320304, 355 pages                 f 50,-

O A macro model for a small economy, summary  f 17,-

O A macro model for Malta with disk, 44 pages f 22,-

