Non-agricultural seasonal and temporary migrant workers in urban and rural Scotland: report

Survey estimates

A mixed-mode survey of employers was conducted between the 1^st of July and the 2^nd of August 2019, yielding 1067 completed interviews (177 online and 890 by telephone).

• Overall 40 businesses said that they employ seasonal migrant workers, comprising a prevalence of 3.7%.
• 1,747 seasonal migrant workers were employed in the 40 businesses who reported that they employ any seasonal migrant workers, with a minimum of 1 seasonal worker through to a maximum of 500 seasonal workers, indicating a high level of variance in the data.
• The mean number of seasonal migrant workers is 42 and median is 6
• The standard deviation of the number of seasonal migrant workers is 96

Sectoral profile of seasonal migrant workers

The data indicates that there is a concentration of seasonal migrant workers within the fishing sector, primarily involving seafood processing businesses, with 996 reported employees within just four firms. This is consistent with Scottish Government analysis into employment patterns within the seafood processing sector which has identified a small proportion of seasonal staff employed within the sector – the report also identifies that 58% of workers are from EEA countries^[21].

On the banking, finance and insurance sector it is worth noting that these businesses comprise recruitment agencies that employ seasonal and temporary workers for hospitality, food processing and cleaning businesses. As the SIC code categories are coded by businesses themselves there is often a lack of consistency in how businesses are categorised using this classification system.

The concentration of seasonal migrant workers within these two sectors indicate a softer finding of the prevalence of seasonal employment within a small number of firms in these sectors. A recruitment agency and shellfish factory were included in the qualitative case studies to explore the use of seasonal migrant labour within these sectors.

One aspect of the data which we found surprising was the low level of response received from the construction sector. As part of our sample design we had assumed that construction may demonstrate seasonal employment patterns as the timing of projects are presumably affected by the weather.

To investigate this finding in more detail, we spoke to some construction businesses that took part in the survey and had said that they do not employ seasonal migrant workers. We found that among construction businesses there was a reliance on oversees employment from the EU8, however, this was on a temporary or freelance basis as opposed to being seasonal patterns of employment. For some construction businesses there was a greater share of local employment – the 2011 census found that only 5% of construction workers were from outside the UK.

Estimates

In using the survey data to produce estimates of the seasonal workforce, we have focused on the distribution, hotel and restaurant sector as this is where there is a greater share of data points for grossing to the population (n=21).

To arrive at an estimate of the total number of seasonal migrant workers in the distribution, hotels and restaurants sector, we split our sample of firms in the sector into size bands. We calculated the proportion of firms in that size band in our sample where the number of migrant workers is non-zero. We then calculated the share of total employment that migrant workers comprise, in each of these size bands. Combining these two pieces of data, alongside total employment in each size band using information from the Annual Population Survey, we were able to gross up our survey estimates to estimate a figure for the number of seasonal migrant workers employed in the sector.

As shown in the table below, it can be seen that 7,100 seasonal migrant workers were employed in the distribution, hotel and restaurant sector.

Moreover by looking at the proportions of those businesses who reported that they do not employ seasonal migrant workers – 388 compared with 21 – we can see there is a low intensity of seasonal employment within the sector.

Aside from the distribution, hotels and restaurants sector, we have been asked to provide an estimate of what our sample implies about the number of non-agricultural seasonal migrant workers across the whole economy. Doing this on the basis of the sample that we have collected through the survey poses a number of challenges.

Firstly, we have very small sample sizes in some sectors, and when we consider sector and size band, some cells are empty. This is why in our main results we focus exclusively on the sector (distribution, hotels and restaurants) where we have the largest and most robust sample. In grossing up our survey data to a population figure, we need to have a large sample of firms in different sectors and across each different size bands. This enables us to capture differences between seasonal worker patterns in each sector and across different types of businesses within each sector. This is not possible with the sample that we have.

The only thing that is possible is to disregard sectoral differences and focus on employment by size band. This, in turn, is problematic for a number of reasons. Principally, that it equates a firm in the manufacturing sector with a similar sized firm in the distribution, hotels and restaurants sector. We have already identified that there are sectoral differences in how respondents define seasonal migrant workers, and hence our results might reflect this. For example, we know by speaking to those who responded to the survey within the construction sector that they tend to use the term temporary workers as opposed to seasonal workers.

In addition, our sample for the intensity of seasonal migrant worker employment within firms is based on only 40 data points across the whole economy. Grossing these up as representative of the entire population of firms in each size band is fraught with problems. In addition, some sectors are more prevalent in our sample of firms in each size band than others, this again suggests that the aggregate number produced by this process is unlikely to be informative. Nevertheless, as it is possible to produce such a figure, the number of non-agricultural seasonal migrant workers in the economy is 51,400.