**2. Literature Review**

Gibrat's Law has been an inspiration for many international publications (Daunfeldt and Halvarsson 2015) and many different methods and approaches have been applied. Mansfield (1962) applied Gibrat's Law in different ways. First, he tested if smaller firms were more likely to leave the market than larger firms. Then, based on economic theory, he investigated whether the companies had to pass a certain scale level, at which production exceeds the minimum efficient scale (MES) level, before Gibrat's Law holds. That is, it is possible that there is a threshold size at which a firm's growth pattern changes. Several other researchers have reported similar results, with the general conclusion being that larger firms grow independently of their size, as for the largest U.S. companies (Hymer and Pashigian 1962; Simon and Bonini 1958). The literature yields a mixed picture for industrial firms, as others show that this does not hold for small- and medium-sized enterprises (SMEs) (Becchetti and Trovato 2002; Hart and Oulton 1999; Fotopoulos and Louri 2004), whereas others show that the LPE holds for entire sectors (Buckley et al. 1984; Hymer and Pashigian 1962; Lensink et al. 2005; Simon and Bonini 1958). Most studies that reject Gibrat's Law show that the sector has a mean reverting tendency, meaning that smaller firms grows faster than larger firms (Almus 2000; Bartoloni et al. 2020; Daunfeldt and Elert 2013; Yadav et al. 2020). According to Jurado et al. (2021), Gibrat's Law applies to large capital-intensive companies that use advanced technology, which is taken to the extreme in other studies who suggest that large firms grow faster than their smaller competitors (Mukhopadhyay and AmirKhalkhali 2010). Finally, there are articles that did not find any hold for Gibrat's Law (Lotti et al. 2001).

Previous research has shown that Gibrat's Law applies to campsites based on data from the Netherlands (Audretsch et al. 2004) and Italy (Piergiovanni et al. 2003), meaning that it cannot be rejected that growth is independent of size. Furthermore, several authors have argued that Gibrat's Law applies to the service sector to a far greater extent than to manufacturing (Audretsch et al. 2004). The articles that failed in rejecting Gibrat's Law for campsites had rather short time dimensions of five years; however, this was in contrast to our ten years studied: from 2010 to 2019. This longer time dimension allows us to use the GMM framework to estimate the parameters of interest, whereas previous analyses of camping sites have used the less advanced OLS framework suggested by Chesher (1979). The choice of estimator is crucial when working with dynamic models using short panel

data, which is why we apply three different estimators that each have their strengths and weaknesses. Through this, we hope that we have attained results that are reliable, so that we can link them to economic theory with certainty.

If there is a minimum size a firm must obtain for survival (the MES), negative growth for small firms may result in deficits, which could, in the long term, lead to the closure of these firms (Audretsch et al. 2004). Mansfield (1962) reported that small firms have relatively higher death rates, but those that survive seem to have higher variation and grow faster than the big firms. He also noticed that firms with successful innovators grow twice as fast as others.

There is a substantial difference between the manufacturing and service industry (Audretsch et al. 2004). In the manufacturing industry, which depends heavily on capital as input, being small is obviously a drawback due to the economies of scale. If production exceeds the MES, the possibilities for profitable operation are far better. This might not be the case for the service industry, since the production is far less capital intensive. Consequently, there are less sunk costs, and economies of scale do not play as much of a key role as for manufacturing. This may explain why there are many small businesses in the service sector. There is a high proportion of family owned units in the Netherlands, and they often do not have ambitions to expand further. The same trend is also found in Italy, where a high proportion of companies in the hospitality sector have fewer than five employees (Piergiovanni et al. 2003). In Norway, the campsites are even smaller.

In a study of Italian hospitality industry (cafeterias, restaurants, cafes, campsites, and hotels), only cafeterias and campsites did not reject Gibrat's Law (Piergiovanni et al. 2003). An analysis based on Dutch data gave the same result for campsites (Audretsch et al. 2004). They state that growth in this sector is independent of firm size. For the four other sub-sectors within the field of hospitality, Gibrat's Law failed to hold. Park and Kim (2010) rejected Gibrat's Law for the restaurant industry, whereas Host et al. (2018) reported that the average growth of firms in the Croatian tourism sector was independent of their size. However, Ivandi´c (2015) did not confirm Gibrat's Law in a study of the hotel sector in Croatia, instead showing that the hotels tend to revert to a certain mean: smaller firms grew faster than the bigger ones. The growth was also shown to depend on ownership, where publicly owned companies had lower growth than privately owned ones.
