4.2. Independent Variables
Due to the lack of systematic and periodic information, reliable statistics and methodologies in Ecuador that allow for the establishment of the identification of creative industries and people working in these sectors, it was decided to take as a reference the occupational data by sector of activity that is provided by the National Employment, Unemployment, and Underemployment Survey (ENEMDU) for 2009 and 2014 [
60,
61]. The people who, as of December of each year, are categorized within the “population with employment” category were taken as a common element. According to the National Statistics and Census Institute (INEC) of Ecuador, the employed population is made up of people aged 15 and over, who during the reference week, were engaged in some activity to produce goods or provide services in exchange for remuneration or benefits.
The creative jobs and occupations of the selected population are part of or are related to the productive chain in the following areas: architecture, handicrafts, visual arts, audiovisual, dance, design, education, photography, music, computer media, and theatre. All this corresponds to one of the definitions of the creative industry pointed out by Caves [
62], as that which produces goods and services that we commonly associate with culture, art, or simply entertainment, producing also creative elements of a strongly symbolic nature. This also matches the definition provided by UNESCO [
63], which mentions that creative industries involve a broader set of activities that include cultural industries plus all artistic or cultural production, either performances or individually produced goods. Creative industries are those in which the product or service contains a substantial artistic or creative element and includes sectors such as architecture and advertising. In this study, 106 subcategories of creative occupations extracted from the International Standard Industrial Classification of All Economic Activities (ISIC) have been considered.
In the first place, “occupations in creative industries” of the National Employment, Unemployment, and Underemployment Surveys-ENEMDU in the period 2010–2017, under the International Standard Industrial Classification of All Economic Activities (ISIC 4.0), were standardised with their equivalent in the National Classification of Economic Activities for Ecuador. After that, the employment situation was analysed for the different regions (provinces) of the country. It is important to mention that the category and subcategory standards of ISIC were used to identify the economic sector, which is the defined creative professions that are part of subcategories. However, for the years of study (2009 and 2014), for some of the occupations, there is no disaggregation at the subcategory level, so in these cases, the data of the corresponding upper immediate category data were used.
According to the data obtained, people with creative jobs and occupations in Ecuador tripled during the five-year period of analysis in Ecuador, going from 271,381 in 2009 to 646,452 in 2014. This issue corresponds to Luzardo, De Jesús, and Pérez [
64], who in a report to the Inter-American Development Bank (IDB), point out that these types of enterprises and businesses have generated thousands of jobs in recent years in Latin America and the Caribbean. In addition, the significant contribution of this sector is shown by exacerbating its importance in the labour market, since if in 2009, 4 out of every 100 employed people performed their main activity in creative areas, up until the year 2014, there were 9 out of 100 people with creative-related jobs, according to data in
Table 2.
Despite the fact that, in the economy as a whole, there is a notable presence of the weight and influence of creative employment, the largest increases in employment participation at regional level are recorded in four provinces of the Ecuadorian highlands: Tungurahua, Imbabura, Chimborazo, and Azuay. The same trend is observed for the case of Pichincha, although its preponderance is lower. These elements seem relevant to initially identifying creativity clusters because four of the five provinces mentioned are located in the northern highlands and share administrative boundaries.
Regional multipliers of creative employment are defined as the number of jobs generated per million dollars of gross value added. When estimating these indicators, a very strong structural change is observed that acquires a countertrend regarding the contribution of the creative economy to employment. A generalised decrease is indeed observed, while in 2009, this multiplier was 247 per million nonoil GVA, in 2014, there was a decrease that reached 133 jobs. This would allow one to infer that the lack of productive chains of these sectors and the high sensitivity of this sector to the economic problems of the country are due to the fall in state revenues caused by the fall in crude oil prices in 2014.
If the behaviour of creative employment multipliers is considered regionally, unequal behaviour is observed (
Figure 1).
The multiplier values were higher in the provinces with lower capital intensity in 2009, which were Pastaza, Esmeraldas, Napo, and Los Ríos. Only two of the regions are maintained in both periods—Esmeraldas and Los Ríos, in addition to Sucumbíos as a result of being one of the regions where the creative employment multiplier increased. On the other hand, Sucumbíos, Zamora Chinchipe, and Santa Elena, regions of lower capital intensity, are the only three regions that increased their multipliers.
The regions with lower creative multiplier values were provinces of the north-central highlands of Ecuador—Tungurahua, Chimborazo, and Imbabura—a finding that was not considered at the beginning of this study approach because footwear, leather, clothing, textiles, clothing, and furniture industries are important in these provinces, and they are also noted for their inhabitants’ manual skills (crafts), that is, creative activities and professions.
As a second variable of the factors that, from a theoretical point of view, have been proposed as determinants of regional economic growth and as a proxy of the capital factor is the amount of the total effective tax collection at the regional level. This variable has been consulted in the general statistics of tax collection of the Internal Revenue Service (IRS).
According to National Information System [
65], the tax contribution consists of direct taxes (income tax levied, environmental vehicle pollution tax, motor vehicle tax, tax on currency outflows, taxes on foreign assets, RISE, royalties, mining conservation patent and profits, rural lands, contribution to comprehensive cancer care, deferred interest payment, tax fines) and indirect taxes (value added tax, special consumption tax, plastic bottles redeemable tax). Tax collection is a consistent indicator to measure the dynamics and structure of capital accumulation of regional systems (
Table 3).
For a better description of tax collections at the regional level, the collection per inhabitant will be used as an indicator. According to the information in
Table 3, at the per capita level, the Pichincha, Guayas, and Azuay regions, where the productive means and the population of the country are concentrated, will contribute the most to the tax rate collection. The tax contribution per inhabitant in its territorial dimension has a great variance, while in Pichincha and Guayas it exceeds
$700, in Santa Elena and Bolívar it does not reach
$70, that is, the difference is of 10 times.
Regarding an increase in tax collection, at national level it doubled. In 2007, the average collection per inhabitant was $177, and after 5 years, it was $302. Both in absolute terms and in proportion, the greatest increase takes place in the province of Zamora Chinchipe as a result of the increase in mining activity in this province. For the rest of the country, the behaviour of this indicator shows a varied typology, where tax collection increases in monetary units do not necessarily have a direct relationship with the increase in relative terms.
4.3. Regressions
Given the data of the variables that make up the proposed model, in the first place, the Spearman correlation coefficient was calculated to check the existence of a relationship of the independent variables with respect to the dependent one. The results affirm that in the case of Ecuador, creative employment predicts economic growth better in 2014 than in 2009. This in turn allows one to argue that it is possible to perform regressions and estimation of parameters between the aforementioned variables that enable the inference of information and the processing of the interrelation and interaction in the territories (
Table 4).
For the logarithmic model proposed, in the economy of Ecuador in 2009, the production elasticities regarding creative employment and tax collections were 0.485 and 0.304, respectively. That is to say that in the provinces of Ecuador, maintaining input taxes constant, an increase of 1% in the creative employment input caused on average, an increase of about 0.48% in production. Similarly, keeping the employment variable constant, an increase of 1% in the capital input generated, on average, an increase of about 0.3% in production. The result of the sum of the two production elasticities is 0.78, which is proof of diminishing returns of scale. In 2014, the influence of independent variables on regional growth is greater, with decreasing returns to scale with a 0.87 elasticity of the two productive factors used (
Table 5).
In the results of the first model of grouped ordinary least squares (OLS), the coefficients of the variables are very significant statistically, with a high R2 value, and coincide with the previous specifications approached in the study. In order to include a dimension that addresses space and time, a balanced panel regression was carried out. In this case, the same territorial unit (province) was studied for the years 2009 and 2014 as a whole, which allowed for the analysis of a greater quantity of data, with higher variability and greater efficiency.
One of the main problems and limitations of OLS is that it hides the heterogeneity (individuality or singularity) that could arise in the provinces of Ecuador, that is, it does not allow one to distinguish if the influence of creative employment and tax collections on provincial growth is the same over time among the different regions. To solve this problem, the panel model of fixed effects and random effects was estimated, the results of which are shown below. For both cases, the F test of individual effects was carried out, which allowed us to reject the null hypothesis that the individual effects are equal to 0 (Prob > F = 0.000), which justifies carrying out an analysis with individual effects. Thus, the grouped OLS model was rejected, since the panel model explains the behaviour and interaction of the variables better.
If the results of the regressions of fixed effects and random effects are compared, a considerable difference between them is observed. Therefore, after performing the Hausman test that determines a χ
2 of 0.12 and a prob > equal to 0.9421 (greater than 0.05), the estimator for fixed effects was selected (
Table 6).