Variable definitions

VN-1: Birth Rate.
The User can change this variable

This is the number of births per thousand women aged 15 to 49 in the current year. As a rule, this rate is declining for the younger countries and relatively static in the range of 35 to 45 per thousand for older countries.

Its initial impact is purely on the number of 0—to 14-year-old children (VN-40) for the first 15 years, impacting household size (VN-6)- but NOT the number of households.

It does not affect the size of the working-age population until 15 years later when those children reach ‘working age’ (defined as 15 to 64/74 years).

Its impact on the labour force (VN-20) is a function of the propensity to be employed in those future years.

The female portion of these births also affects the number of women of childbearing (VN-3) age after 15 years – and then ultimately the number of births in that year (VN-4)

Note it does not impact the number of households (VN-7). Newborns are usually an addition to an existing household.

VN-2: Migration rate as a per cent of the total population.
 The User can change this variable.

This is expressed as a percentage of the previous year’s total population, typically within + or – 1%. A negative is emigration, and a positive is immigration. It is assumed they come/leave as families, affecting the age range of 0 to 64.

It should be noted that this variable can change significantly from year to year due to government action or the economic environment.

This variable is crucial as it impacts several other variables. They are:

• • The number of people by age (VN-40 to VN-43)
  • The number of women of childbearing age (VN-3)

Those variables, in turn, impact the following variables in the same and subsequent years:

• • The number of persons of working age and hence the number employed) after allowing for propensity to be employed)
  • Total population
  • Number of households

The total number of migrants each year is held in Variable VN-8.

Vn-3: Women aged 15-49

This changes from the base scenario if:

• • There is a change (increase or decrease) in total migration (VN-51) – because a portion of migrants will be females aged 15-49
  • There is a change in total births (VN-4) from the base scenario due to a change in the birth rate (Vn-1). The female proportion of births after the base year is added to the passive forecast of females aged 15-49 years 15 years later and onward.

This variable is reported in 000’s.

VN-4: Total Births

This is simply the number of women aged 15 to 49 years (Vn-3) in the specific (forecast year) multiplied by the birth rate (VN-1) for that year.

However, note that the model uses 5-year age groups for the calculation. This is necessary as the birth rate does vary by the age of women—under 34-year-olds are significantly more likely to have a child in any one year than those over 34.

The model has the long-term birth rate trend for each age group.  But these will be varied proportionately if the user alters the average rate (VN-1)

Total births are reported in 000’s

VN-9: Males aged 15 to 64 years

This subset of males in variables VN-41 to VN-43 is termed working-age males. The population’s overall age profile determines its size.

It is reported in 000’s.

VN-10: Males aged 65 to 74 years

This is a subset of males aged 65 to 74. This is important as an increasing proportion of these males are still in full-time employment (25% in some countries) and are of growing importance to the overall workforce size. The expectation is that this age segment will increase proportionately to the total population over time as the population ages.

It is reported in 000’s.

VN-11: Females aged 15 to 64 years

This subset of females in variables VN-41 to VN-43 is termed working-age females.

It is reported in 000’s.

VN-12: Females aged 65 to 74 years

This is a subset of females aged 65 to 74. Again, it is an important group as it increasingly contributes to the overall size of the working-age population.

It is reported in 000’s.

VN-13  Total Working Age

This is the sum of variables Vn-9 to VN-11

It is reported in 000’s.

Propensity to be Employed (VN14 to VN-17)
 The user can change these variables.

This is the probability of a person in the defined age range being in employment.  It does not include those seeking work but currently is unable to find it (the standard definition of unemployed)

It is these four variables multiplied with the previous variables (9 to 12 – working age) that determine the size of the employed labour force in each year,

VN-14: Propensity to be Employed – Males 15-64:
The User can change this variable.

This is the proportion of all 15- —to 64-year-old males in full-time employment (typically 24 hours or more per week).

This varies significantly by country and is partially correlated to the level of education.  It is also a function of the global economy and the extent of government expenditure (creating employment opportunities).

The user can vary the percentage. The recommended maximum is 90%—it is rarely above that. It can drop to as low as 35% in less well-educated and poorer countries. Be aware that changes in this variable tend to be gradual over time. However, a sudden economic downturn can quickly result in high unemployment for a short time.

This age group typically has the highest propensity to be employed of any age/gender group.

This variable, multiplied with the number of males aged 15 to 49 (VN-9) gives the expected number of employed males aged 15 to 64. (VN-18)

This variable is expressed as a percentage (%) of all males aged 15 to 64.

VN-15: Propensity to be Employed – Males 65-74:
The User can change this variable.

In countries where education standards are high, and life expectancy is into the eighties, the propensity of this age group to be still working is increasing, and in some countries, it is already over 20% of such persons. We recommend that for countries with a GDP per capita over US$20,000, this variable be set at least at 5%. However, given current trends in Japan, which probably lead to this trend, it is unlikely to exceed 35%.

This variable, multiplied by the number of males aged 65 to 74 (VN-10) determines the number of employed 64 to 74-year-olds (VN-19)

This variable is expressed as a percentage (%) of all males aged 15 to 64.

VN-16: Propensity to be Employed – Females 15-64:
The User can change this variable.

This is the female equivalent of VN-14. In many countries, this proportion is still increasing as attitudes change about the role of the female. It is starting to plateau in the more affluent and older countries but probably has not peaked. It is perhaps best to consider that it will continue to increase—albeit slowly in the older, more affluent countries.

In no country is the percentage of females aged 15 to 64 employed higher than that for males. It would be unwise to set a scenario where it did. However, where the employment rate is low, it is generally much lower for females than for males.

This variable, together with the number of females aged 15 to 64 (VN-11), determines the number of employed females aged 15 to 64 (VN-20).

This variable is expressed as a percentage (%) of all males aged 15 to 64 years.

VN-17: Propensity to be Employed – Females 65-74:
The User can change this variable.

This is the proportion of females aged 65 to 74 in full-time employment.  As a proportion, it is currently less than that of the equivalent male age group.  However, it has also been increasing over time, and that may accelerate in future.  Again, it is considered unlikely to exceed the percentage for males.

This variable, multiplied by the number of females aged 65 to 74  (VN-12) determines the number of employed females aged 65 to 74 (VN-21)

VN-18 to VN-21:  Number of employed persons by age group

The values here are determined by

• • applying the propensity to be employed for each age group (VN-12 to VN-15), which the user can change to
  • the number of persons in each gender/age group (VN-8 to VN-11). These values are determined by the trend in the population’s age profile, which is a function of births, deaths, and migration and, therefore, not directly within the user’s control.

VN-22: Total Employed Persons

This is an important variable as the employed labour force is the economy’s ‘engine’. Productivity per worker also counts, but the underlying issue is the number of workers in the economy.

It is derived from the sum of employed persons by age (VN-18 to VN-21), determined by the number of persons in each age/gender segment (VN-9 to VN-12) and their propensity to be employed.  (VN-14 to VN17).

Note that the user can influence this variable by varying the Propensity to be employed (VN-14 to VN-17)

VN-23:  Employed per Household

This is the total employed (VN-22) divided by the number of households (VN-7).  The higher this is, the more wage earners the household has and, typically, the more affluent it is.

This is one reason why household incomes are not growing as fast in the more developed societies: They are older, and there are fewer additional workers in the household. Typically these older households are two or fewer persons.

VN-24:  Percent of Working Age who are Employed

The total employed (VN-22) is divided by the total working-age population (VN-13).

This gives a comparative measure (across countries) of the proportion of the working-age population that is productive.  The higher the percentage, the better.  It is a valuable measure of the economy’s ability to create employment opportunities.  Education is a crucial driver in terms of this.

VN-25:  Percent of Population Employed

This is the total employed (VN-22) divided by the total population (VN-5).

This gives a comparative measure (across countries) of the proportion of the population that is productive.  The higher the per cent, the better.

VN-26:  Dependency Ratio

While this is simply a derivative of the total population (VN-5) and total employed (VN-22), it is an important variable explaining the society’s capabilities.  It is calculated by subtracting the total employed population from the total population to get the total non-working population and dividing that by the total working population.

If it is below 1, there are more working people than non-working, and society has a lot of spending power per capita.  However, if it is substantially over 1 (say 1.5), then every worker is supporting themselves and 1.5 other persons.  – a total of 2.5 persons supported per wage earner.  That impacts the household’s spending power compared to a society where the workers are supporting themselves and perhaps less than one other person.

VN-27:  Fixed Capital Investment (FCI) as a per cent of GDP of the previous year. 

The percentage of GDP (previous year) that goes into Fixed Capital Investment varies significantly across countries and over time.  The global norm is 27%.  The variation tends to be influenced by the need to stimulate the economy (increased FCI) or excess government debt (reduced FCI) or to maintain adequate, if not complete, employment (Typically increased FCI unless there is a labour shortage).

Not all fixed capital expenditures are government expenditures—the private sector can account for a significant proportion. Still, the government can influence this through investment incentives, such as FCI being tax deductible.

VN-28:  Fixed Capital Investment for the year

This is determined by multiplying the level of Fixed Capital Investment per worker (VN-29 user-determined) by the number of employed persons (VN-22).

The model is structured this way as analysis across the 118 countries covered by Global Demographics Ltd demonstrates that the size of the workforce is a significant determinant of the level of FCI that takes place.  Obviously, governments can vary from that relationship, but this can have consequences.  If it is too high relative to the labour supply, it causes wage inflation and perhaps general inflation.  If it is too low, then unemployment can be a possible outcome.

VN-29:  Fixed Capital Investment per worker for the current year
The User can change this variable.

This is simply the country’s Fixed Capital Investment (VN-28) for that year divided by the number of employed persons for the same year (VN-22).

However, this variable can be set by the user (rather than using the default forecast value based on historical trends).  It is generally relatively stable in value and trend but can be increased if the share of GDP spent on FCI starts to decline due to the total GDP growing at a faster rate than FCI per worker due to increased productivity.

While it is possible that such additional revenue could be allocated to FCI, be aware that with increased affluence, government spending typically moves to social services rather than FCI, maintaining rather than increasing the absolute amount of FCI per worker.

It is recommended that you experiment with this variable as it does impact GDP per worker (Gross productivity VN-33) and, hence, ultimately total GDP.

Also, be aware that you can override the estimated normative impact of this variable on productivity by directly altering productivity per worker (VN-33).  See more comments on this below.

VN-30:  Accumulated Fixed Capital Investment per worker

This is an important variable as it is a proxy of the total facilities available to the worker.  The hypothesis (supported by data from 118 countries) is that the higher this is, the more productive the worker is (but the lower the proportion of GDP per worker paid out in wages).

As Total Fixed Capital Investment can vary a lot year on year depending on government policy), it is necessary to take a longer-term view of the capital equipment available to the worker. This is achieved by summing the FCI for the previous ten years,  depreciated at 10% per annum.

The user can influence this variable by altering the FCI per worker (VN-29 above), but be aware it is only 10/55ths of the Accumulated FCI per worker in any one year – so the impact is not as great as one might expect.  Legacy investment levels have to be taken into account.

VN-31:  Education Index.
The User can change this variable.

In many respects, this is a population variable, but its impact is economic—that is, it significantly impacts the workforce’s productivity. It achieves this in two ways.

• • First, by definition, the worker has a higher skill level and typically can use a greater range of facilities (e.g., computers) to lift their output. However, note that the relationship between education and productivity is not linear. The higher the education standard, the less the incremental impact on productivity.
  • Second, education appears to attract investment. The higher the education index value, the greater the FCI per worker (historically) and the greater the productivity. Do note that causality is assumed but not proven.

Do appreciate that education is an influential variable. Over the medium to long term, it also influences (albeit passively) a country’s birth rate (VN-1).  Almost universally, the higher the education standard, the lower the birth rate.  However, other factors, such as religious norms, can also influence that relationship.

The Education Index is a weighted measure of the education profile of the adult population (15 years and above) and a proxy measure of the education level of the employed population. The higher the index, the better educated the workforce will be, especially as the better educated are more likely to be employed.

It is constructed as follows:  It is the sum of:

• • % with primary or less * 1
  • % with Lower secondary * 2
  • % with upper secondary * 3
  • % with tertiary or vocational * 4

The higher the index value, the better educated the labour force. Tested across 118 countries, this positively correlates with worker productivity (and Fixed Capital Investment per worker).

The user can change this variable.  However, it is not recommended unless the user is aware of a particular boost in education spending by a country (such as India in the year 2000).  When making modifications because of this, be mindful that the impact of this change in the availability of education will not reach the labour force (compared to the school-age population) until at least ten years in the future.  That is when the better-educated children reach working age and enter the labour force.  So, for example, India’s improved availability of education probably will not be reflected in the adult population (and workforce) until 2035 at the earliest.  So, that is the earliest year the user should change the index value.

In contrast, note that the education profile of immigrants is often different from that of the resident population.  Typically, immigrants are less well-educated and lower the overall index value.  So, if there is a significant level of immigration, the education index may not increase as fast as otherwise expected for that country.  But also note the opposite is true.  Emigrants tend to be better skilled and can move to other (higher paying) countries, thereby lowering the home country’s education profile.

Productivity Equation

This is not a variable but an equation.  It intervenes between the combination of Education (Vn-31), Accumulated Fixed Capital Investment per worker (VN-30) and GDP per worker (VN -33).

The model uses the average percentage change in the previous year’s education index and accumulated fixed capital investment per worker as the single explanatory variable.  The dependent variable is the percentage change in GDP per worker (Productivity measure).   Change, rather than absolute values, is used as there is no consistency across countries in terms of absolute values.  However, there is a consistent relationship between changes in the two driver variables and the dependent variable, both over time and across countries.

The equation model states that the change in GDP per worker is a function of the average change in the two variables that potentially influence productivity: skill (education) and facilities (FCI) per worker. The equation is solved separately for each country and tends to have a well-explained variance, typically over 0.6.

VN-33:  GDP per worker (Gross Productivity per worker)
The User can change this variable.

This is an important variable as it significantly determines the overall size of the economy, the income level of the individual household, and, hence, lifestyles.

This is simply the Gross Domestic Product per worker – obtained from applying the equation of the relationship between changes in education and accumulated Fixed Capital Investment per worker and the change output per worker as measured by GDP per worker.  The previous year’s GDP per worker is the base used.  This is explained in more detail above.

However, it has been made a user-changeable variable as there are instances where the relationship, as demonstrated in the equation, does not hold.  For example, a sudden drop in exports lowers the overall productivity per worker – even though there has been no decline in education or Accumulated Fixed Capital Investment. (the enablers)

The values shown for this variable in the first instance are the default relationship.  If other factors have changed, the user can override the expected normative values with your assumed values.  For example, productivity drops by 10% from what it otherwise would have been.

You can enter any value other than zero or below.  However, variations in the previous years can be seen as indicators of the likely limits of such changes.  Also, be aware that reducing employment rather than productivity per worker might be better. (same effect on total GDP).

VN-34: Total Gross Domestic Product

This is GDP per worker (VN-33) multiplied by the number of employed persons (VN-22)

VN-35: Gross Domestic Product per capita

This is the total GDP (VN-34) divided by the number of persons in the population (VN-5).

This is a more important variable than total GDP (VN-34) as it shows how individual prosperity changes.

VN-36:  Wage Ratio.
The User can change this variable.

This is an important variable as it directly impacts household incomes and employers’ profitability (public or private).

It is the proportion of GDP per worker paid out in wages.  Data to 2022 is the actual relationship.  The forecast uses an equation where the proportion paid out is a function of:

• • The proportion of the working-age population in the work – a measure of how tight the labour market is (VN-24)
  • The accumulated Fixed Capital Investment level per worker is the extent to which capital replaces/enhances labour. (VN-30)
• The skill level of the worker (education index VN-31)

Analysis across 118 countries has found that the wage ratio:

• • Increases as employment as a percentage of the population increases, presumably as a result of greater competition for scarce labour and
  • Decreases the more significantly the accumulated Fixed Capital Investment per worker increases, presumably because the cost of capital involved needs to be paid

The model provides the expected state, but the user can override this and raise or lower it accordingly.  Note that it impacts wages, household incomes and consumer spending.   However, it can also impact investment (FCI).  If wages increase as a proportion of total productivity, there is less to invest, and productivity per worker grows slower. (VN-33)

The model does not have a feedback loop for this as it has proved difficult to quantify. The user can experiment with the effect by first (say) increasing wages as a percentage of GDP per worker and then, three years later, slowing FCI per worker and seeing the impact.

VN-37:  Average Wage per Worker

This is the wage ratio (VN-36) applied to the GDP per worker (VN-33)

N-38:  Average Household Income.

This is simply the average wage per worker (VN-37) multiplied by the average number of employees per household (VN-23).

These two variables are determined by Productivity per worker and propensity to be employed, which occur earlier in the model.

VN-39: Average Household Income per capita

The average household income (VN-38)is divided by the average household size (persons per household VN-6).

VN-40: Average Propensity to Spend
The User can change this variable.

This is the average proportion of gross income (including taxes and savings) spent on consumption.

The model shows the country’s historical levels up to 2023 (the latest actual) and projects forward based on that trend relative to gross household income.

Analysis across 118 countries shows that the propensity to spend has a consistent and logical pattern.  That is, initially, when incomes are low, the tendency to save is very low and remains so for quite a wide range – typically about US$5,000 pa.  Then, incomes get to the point where the household has enough of the basics to survive and the propensity to save increases steadily after that.  The rate varies depending on the future security of income.

The user can modify the proportion spent, but be aware that it should not exceed 100%  (it is an average for all households, not just low-income who can spend more if receiving other income support).  Care should also be taken to ensure that the total spending of all households does not result in a private consumption expenditure component of GDP becoming excessive.  (see comment on VN-44 below)

VN-41: Average Household Expenditure

This is the average household Income (VN-38) multiplied by the average household’s propensity to spend (VN-40)

VN-42: Average Household Expenditure per cap

This is the average expenditure per household (VN-41) divided by the average household size (VN-6)

VN-43: Total Private Consumption Expenditure

This is the average expenditure per household (VN-41) multiplied by the number of households (VN-7) and divided by .93.

The adjustment of 0.93 allows for expenditures by charities and other social services on households. Such expenditures are typically included in the nationally reported private consumption expenditure, and while the exact proportion is not generally published, it averages 7% where published.

VN-44: Private Consumption Expenditure as a % of Gross Domestic Product

This is a sanity check.  Total Private Consumption Expenditure is a percentage of the same year’s total Gross Domestic Product. There is no specific ‘norm’ for what this percentage should be, but it is unusual for it to be below 50%.

As a rule, it is low in more developed countries with higher taxes, so government (public) spending becomes significant. In less developed/affluent countries, private consumption expenditures are much higher. Eastern Europe runs at 60%, developing Asia is around 70%, and South America is close to 70%.

VN-44: Private Consumption Expenditure as a % of Gross Domestic Product

N-70:  Average Household Income.

This is the average gross income of all households. It is the same as Variable VN-38

This is reported as actual US$ income per annum.

VN-71:  Median Household Income as a Percentage of the Mean.
The user can change this variable.

This is the median as a percentage of the mean. By definition, it will be at or below 100%.

The lower the value, the more uneven the country’s income distribution by household. For example, if the median is 66% of the mean, two-thirds of households have an income lower than the reported median (VN-71). This means the other third of households have significantly higher incomes than the average (or median), accounting for half the total income of all households.

Generally, this variance increases as society becomes more affluent, as there are always some households with very low or no income, which drags the median down relative to the mean.