PKF survey shows early school-leavers exceed EU average

PKF funded a scientific survey among females of different age groups to test the activity rate in the workforce and why Malta is trailing behind the EU28 average – put simply the survey found a number of interesting factors that need to be addressed in order to improve productivity and aim for a higher GDP level.

Examining the years from 2005 to 2014, one finds that the overall female activity rate has increased from year to year. From 36.4% in 2005, the rate has increased to 52.1% in 2014, however although there was a gradual increase over past years, the 2014 rate is still well behind the EU 2014 average of 66.5%.

Apart from quoting the overall activity rate, the National Statistics Office gives the rate for three different age groups, which are 15-24, 25-54, and 55-64.

For the group pertaining to the smallest ages, the rate goes up and down between 47.7% and 51.7% over the years considered, as can be seen in two graphs that follow. This younger age group registers the least activity gender gap, that is, the least difference in the employment rate between males and females.

Furthermore, the 2014 female activity rate for this age group reached 51.7%, which not surprisingly exceeds the EU 2014 average of the same age group by 12.8 percentage points.

This excess is only visible in this age group of 15-24 and allegedly reflects a negative trend, meaning that more females are at work due to a high early school-leaving rate.

For the second age group, that of 25-54, the female activity rate has increased considerably from year to year, being 38.7% in 2005 and increasing to 63.3% in 2014. Unlike the previous age group, the female activity rate in this case is less than the EU 2014 average of the same group, being 79.5%. 

From the review of literature and the discussions made with key stakeholders, PKF concluded that the variables which most influence the female activity rate pertains to education, birth and marriage rates, inflation and the introduction of free childcare centres.

The education variable was divided into other variables, according to the age group being analysed. Furthermore, to undertake regression analysis based on data so collected, we assumed both the crude birth and marriage rates. All this data was obtained from the NSO and Eurostat databases for the 10 years 2005 to 2014.

The crude marriage rate for 2014 was not found, thus to replace this missing value, mean imputation was used. For the casual reader, we note a major limitation of regression models is the fact that due to time constraints and other logistical limitations there was a limited range of data entries collected and so the exact effect of the explanatory variables on the female activity rates may not be fully depicted.

Additionally, the predictor variables considered are only those which data pertaining to them is available for the public and which are believed to have an effect on the female activity rate.

Apart from these variables, however, there may be others which are more influential but these are either unknown or cannot be quantified and represented by a statistical variable.

For example, factors which might considerably influence the females’ decision to start work are ingrained social habits, attitudes and traditions, where females abide by the unwritten rule that they are the ones who take care of families and in the process ought not to commit themselves to careers that distract them from fulfilling this aim.

So in this social habitat, females may not wish to participate in ambitious careers even when the most appealing employment incentives are initiated, and naturally this social factor cannot be catered for in the statistical models. A consolation is that there are effective employment incentives which in Malta proved very successful since more females than males graduate each year at university, so under utilization of this intellectual capital means wasted millions spent on education each year.

New incentives such as child care centres (apart from many others introduced earlier) might have had an effect on the female employment rates; however, these will not be catered for in the regression models. The main reason for not including these factors is that there are many incentives, and consequently, the number of parameters to be estimated in the regression models will exceed the number of data entries.

Starting with the 15-24 age group, the education substrata is influenced by the Early School Leavers and the Youth Educational Attainment variables.

The above equation implies that when the free childcare centres are available, the female activity rate is 3.203 percentage points more than when they are not available, keeping the other effects fixed.

This result implies that the free childcare centres incentive has succeeded in a relatively short timespan to positively affect the activity rate. Furthermore, the model reveals that with a one percentage point increase in the early school leavers’ rate, the female activity rate is increased by 0.197 percentage points.

Moving on to the next 25-54 age group, the education cohort consists of three namely – the Youth Educational Attainment, Lifelong Learning and Tertiary Educational Attainment variables.

The first variable is the same as the one included in the 15-24 age group regression model. This was implied since the educational attainments made by females aged between 20 and 24 might also influence the overall activity rate of the females aged between 25 and 54.

The Lifelong Learning variable represents the percentage of all those females aged between 25 and 64 who involve themselves in courses, training and conferences or participate in regular education. On the other hand, the Tertiary Educational Attainment variable represents the percentage of all female individuals aged between 30 and 34 years who have achieved at least tertiary level of education. Finally, we discovered how the female activity rate of the 55-64 age group has increased, albeit slowly from 12.7% in 2005 to reach 20.6% in 2014. However this age group’s rate is the most distant from the corresponding EU average of 48.40%.

This is regretted and could possibly be due to a low level of computer literacy and lack of participation in lifelong education. No survey is complete without the need to check the activity rate for young females with children up to 16 years of age, so in this instance the primary objective was to establish main reasons for unemployment as well as particular circumstances which may lead to women leaving the labour market.

Such surveys were carried out in the form of face-to-face interviews in prominent places both in Valletta and Sliema, with the majority of the responses obtained in July this year from an electronic version which was circulated on Facebook for 10 days.

In conclusion, there is a lot that needs to be done, such as changing the school opening hours to solve the problems of mothers and to improve education levels of families if we aim to reach the EU activity average of 79% by 2020.

One can say that progress has been registered in some areas yet ideally, a study needs to be jointly sponsored next year by the ministries of education and family welfare to report whether the targets set by 2016 budgetary measures are reached.

Readers may wish to obtain a full copy of the scientific study by writing to Janet Chetcuti (Chief Statistician) at [email protected]