A Macroeconomic Review of the Factors Influencing Fruit Consumption in Romania—The Road towards Sustainability

Abstract

Sustainable development, including the consumption of sustainable food, is an issue that is receiving increasing attention in research and policy construction. Thus, complex policies are being created to address these issues, targeting economic, social, and environmental factors. This study aims to provide a macroeconomic overview of the factors that have the potential to influence fruit consumption at a national level, so that proposals can be offered in order to pursue a more sustainable fruit consumption in Romania. In order to carry out the research, two approaches were used: a bibliometric technique, through which the Scopus and Web of Science publications on the supply and consumption of fruits were analyzed, and an econometric approach, through which some of the macroeconomic factors influencing fruit consumption in Romania were identified and assessed. The study highlights the high interest of worldwide researchers in the field and an upward trend in Romanian fruit consumption.

1. Introduction

There are two main paths of approach to evaluate the importance of fruits in sustainable development. First, starting from the production perspective, organic farming contributes to the sustainability of agriculture by having a low negative impact on the environment [1], providing food security and food safety to consumers [2], while ensuring a fair income to producers [1]. Fruit growing is an important part of the agricultural sector [3]. Second, the consumption of fruits is perceived as healthy [4] and the choice of local, organic, or seasonal fruits responds to sustainable consumption choices [5]. Moreover, most dietary guidelines present the consumption of fruits as an important part of developing a healthy lifestyle, by reducing the risk of developing some chronic disease [6], since they are important sources of vitamins, minerals, micronutrients, and fibers [7], while being available for consumption in different ways, all year-round [8].

Over the years, the consumption of fruits and vegetables has been intensely promoted, including advice from the World Health Organization (WHO) [9], who recommend consuming at least 400 g of fruits and vegetables per day. In addition, the European Union (EU) has provided aid schemes to promote the consumption of fruit and vegetables from an early age, offering children fruit and vegetables in schools as snacks [10], thus emphasizing the importance of developing the habit of consuming them.

Yet, several authors [10,11] and statistical measurements [12] highlight the low consumption of fruits. For example, 36% of Europeans did not eat fruits daily in 2017, and only 42% of Romanians ate fruit on a daily basis [12].

The choice of food is a personal one, considering personal needs, knowledge and priorities, no matter the general recommendations and food policies. In a particular study [13], a series of personal factors that influence food choice have been determined: health, mood, sensory appeal, natural content, weight control, convenience, familiarity, price, and ethical concern. Moreover, it is considered that regionality and cultural differences are important factors for the quantity of fruits and vegetables consumed in different areas [14]. Therefore, policies promoting a standardized fruit and vegetable consumption are not destined to succeed [15].

The political efforts for sustainable development, including promoting sustainable food consumption, appear in various forms, from national dietary guidelines to complex policies that encompass social, economic, and environmental dimensions [16], such as the Agenda 2030 based on the 17 Sustainable Development Goals (SGDs) [17]. Romania, as a member country of the EU, has the responsibility to contribute to achieving the sustainable development targets, including by providing support for its population in understanding and adopting new sustainable consumption ideas [18] and habits [19], including an increase in fruit consumption [20].

The current state of fruit consumption in Romania is a sensitive one, as the European statistics place it considerably under the EU average, on the twenty-third position among the former twenty-eight member states [12]. Even if the absolute quantities of fruits consumed annually have almost doubled in the last thirty years, from 59.5 kg/year/person in 1990 to 110.8 kg/year/person in 2018 [21], that is the equivalent of 303 g/day on average. The quantities are considerably higher in the urban area than in the rural area. The fruits that are mostly consumed in Romania are citrus, with 35% of the yearly consumption in 2018, followed by apples, which comprised more than a quarter of the annual fruit consumption in 2018, and melons with 22% [21]. Local known fruits such as apricots or peaches have almost insignificant percentages in consumption due to low availability and high prices [21].

To be able to provide a targeted national policy for increasing fruit and vegetable consumption, as suggested by other scholars [10], authorities should be aware of the current state of fruit consumption, as well as those macroeconomic factors with the potential to influence it. Therefore, the purpose of this paper is to provide a longitudinal image of macroeconomic factors with the potential of influencing the fruit consumption at national level and to evaluate which of these factors have a positive and negative influence, so as to be able to propose some policy solutions targeted at Romanian consumers.

While several studies concentrate on the importance of adopting a sustainable diet, based on the increased consumption of fruit and vegetables, the current study contributes to the literature by shedding some light on the real possibility of changing the dietary pattern towards a more sustainable diet, considering some macroeconomic factors such as meat consumption, the change in GDP, the change in foreign direct investments, the change in agricultural employment, and temperature change in Romania. This could also support policy changes to stimulate sustainability through fruit consumption in other European countries.

2. Theoretical Background

2.1. Sustainable Development and The Agricultural Sector

Increasing social needs determined by population growth and intensive economic activities have led to the need to achieve a better balance in terms of natural resource use and environmental protection, thus aiming for a long-term vision of sustainable development, considering three important pillars (economic, social, and environmental) [22].

The International Institute for Sustainable Development (IISD) presents the chronology of sustainable development [23]. In 1972, the United Nations (UN) emphasized the relationship between the environment and human development needs, and in 1980 an international strategy (World Conservation Strategy) mentioned the concept of sustainable development from an ecological point of view, which in 1987 was defined in the Brundtland Report as a type of development that responds to current needs without endangering the possibility of the next generations to respond to theirs. In 1992, the UN through the Earth Summit adopted Agenda 21, which aimed to create the UN Commission on Sustainable Development (UNCSD). In 2002, the concept of sustainable development benefited from its own World Summit, where improving the well-being of people’s lives along with the conservation of the natural environment dominated. In 2012, the Rio+ 20 Summit mentioned the need to approach a green economy, so to ensure a sustainable development [23], and in 2015 [17] the 2030 Agenda was created, which presented 17 sustainable development goals (SDGs) that should be met by 2030 in order to achieve sustainability.

The 2030 Agenda [17] has an objective specifically linked to the agri-food sector, SDG 2. It focuses on “ending hunger, achieving food security and improving nutrition and promoting sustainable agriculture” by ensuring access to healthy food for the entire population and increasing the yields, while ensuring sustainable food production systems [24]. Agriculture is one of the strategic economic sectors considered in the achievement of the SDGs, and effective actions for enabling the sustainable development of the agricultural sector must be designed and followed in all three pillars (economic, social, and environmental) [16].

Because SDG 2 emphasizes the sustainable development of the agricultural system as a whole, the Food and Agriculture Organization (FAO) defines sustainable food systems (SFS) as those systems able to provide food security and nutritious food for the entire population, while not hindering the possibility of the future generations of doing the same [25].

The sustainable development of the agricultural sector has been defined by the FAO [26] as being “environmentally non-degrading by conserving the natural resources, economically viable and socially acceptable”. The agricultural sector has always had the capacity to unite economic, social, and environmental aspects through specific actions, which aim at ensuring sufficient and nutritious food. Therefore, the sustainability of the agricultural system is necessary for its development in the current context [27] and the progress towards sustainability may start at the local level, leading to a global integration of the agricultural sector development [28].

2.2. Food Consumption and Sustainable Development

Food consumption is a natural process for every human being; without food people will not be able to survive [29]. The natural resources available at regional levels, the cultural background of each country along with the level of income determine a specific food pattern for each country [30]. The economic progress of the low-income countries generates profound changes in those patterns, as people tend to eat more meat and forget about the traditional food they ate, aiming towards a developed lifestyle, including a new food consumption pattern [31,32,33]. The new food patterns place more pressure on the use of natural resources and significantly contribute to climate change [31,32,33].

Some authors [34,35] point out that a growing category of people have started taking into consideration their individual contribution to diminishing climate change and to sustainable development and adapting their lifestyles so as to diminish their negative impact. Some examples of the individual contribution are, among others, the choice of sustainable fashion [36], the choice of less pollutant transportation methods [37], and even the choice of food [38,39].

The FAO [40] describes sustainable diets as those which are targeting the long-term development of individuals, through:

• Low environmental impacts to protect the natural resources;
• Accessible and affordable food, which is safe and healthy for human consumption;
• Fair trade for ensuring adequate incomes for farmers and the right price for the final consumers.

Some authors [41,42] claim that the main factor in the choice and consumption of sustainable food products is the certification, especially organic farming, or fair trade. Others consider that the choice of a locally produced food product or traditional food [43,44], or a seasonal product [43,45], are characteristics of a sustainable choice, while others emphasize the perceived health considered when choosing a sustainable food product [20]. Therefore, while a commonly accepted definition of sustainable food is still missing [46], there are some common aspects for being sustainable.

In this scenario, increasing fruit consumption responds to multiple sustainability characteristics, for example a part of the research focuses on the health benefits of fruit micronutrients, their high content of water [47] and the important role of the antioxidants found in fruits [48], opposed to observations on the low intake of those micronutrients caused by the low consumption of fresh fruits, leading to weakened immune systems [49].

There are some regional and country variations regarding fruit and vegetable consumption based upon determinant factors, such as economic level differences and socio-cultural influences of the consumers [50]. Furthermore, the socioeconomic status of the consumer is seen as an influencing factor for fruit consumption [51].

Some studies [52,53] point out the importance of diversifying the range of fruits and vegetables in diets, especially in young people, but also the need for boosting the consumption through more attractive and easier to consume products that have developed in the recent years [11], such as:

-

Fresh, ready to eat products that consist of different combinations of fruits and vegetables (premade salads), which are cut, washed and ready to be eaten immediately and anywhere.

-

Freshly cut products that are represented as a healthy snack, such as cut fruits, or various fruits and vegetables cleaned and vacuum-packed that can be prepared faster.

-

Fresh or pasteurized fruit and vegetable juices.

-

Frozen fruits in exact proportions to be turned into milkshakes, ice creams or others.

Some authors [41] indicate that another factor that consumers consider when choosing fruits and vegetables is the quality certification that the products have. Food quality certifications being promoted in accordance with the sustainable goals, help to build consumer confidence in the fruits and vegetables that hold these certifications [54], for example, local, seasonal, and organic certified products [20]. Other authors [43] show that the main five characteristics of fruits and vegetables considered by consumers are origin, seasonality, freshness, local origin, and price.

2.3. Macroeconomic Factors Influencing Food Choice

As several studies show [55,56,57,58], the possibility of changing the dietary pattern does not depend solely on the individual’s will, but on the real options of them doing so. The higher income countries can afford to consider the sustainability-related criteria in choosing food products, unlike countries with lower incomes, where most of the income is spent on food [57]. The economic condition and food prices are important influencing factors in food choice [58]; therefore, changes in the inflation rate are expected to influence fruit consumption [59]. Furthermore, the habituality of eating certain foods, such as fruit, may be obtained through early education, therefore the educational policies should be adapted to include nutritional knowledge, as some authors point out [60]. Other authors [61] demonstrate that factors such as the price of a particular fruit compared to the price of substitute fruits, the consumption per capita of food, the real income of consumers, the general price index of the goods, the technology, and the USD real exchange rate against the currencies of each of the countries considered have significant influences on consumer choice. The general indicator for the level of income, considered in studies on influencing factors of food consumption is the GDP, the results pointing out again that countries with higher GDP have more meat-based diets [31], while the rapid change in dietary patterns puts increased pressure on agricultural supply, as well as on imports and exports of food products [31]. The correlation between GDP level, food supply and food exports is also studied and confirmed by other scholars [62]. Furthermore, the policies intended at reducing the greenhouse gas emissions from agriculture also have an influence on the supply of food, including fruit supply [63].

3. Materials and Methods

The research was conducted by using two approaches.

First, a bibliometric analysis was conducted, considering the articles published on fruit supply and consumption, following the evolution and the general interest in the topic. In order to search for articles published in the Scopus and Web of Science databases, the following search criteria were used (search performed on 31 October 2020): “fruit * supply” and “determined *”, a criterion that generated one result for each database; “fruit * consumption” and “determined *”, generating 37 and 22 results, respectively; “fruit * supply” and “factor *”, with 41 results and 31 results, respectively; “fruit * consumption” and “factor *” with 961 results and 653 results, respectively.

To create the map based on the results obtained through the Scopus platform, with the help of VOSviewer software [64] the selection criterion was defined at five minimum occurrences of a keyword, out of 6697 words. Only 869 met this criterion. In order to generate the map, in the VOSviewer software, a criterion of word occurrence of at least five times was applied. Thus, out of 3331 words only 243 words met this criterion.

Second, an econometric approach was carried out to investigate the macroeconomic factors influencing fruit consumption in Romania during 1990 and 2019 by using the EViews software [65]. The tested variables were chosen by considering all three pillars of sustainability—economic, social, and environmental—and the previous studies in this area [21,31,55,56,57,58,59,60,61,62,63], from the following databases:

(a).

National Institute of Statistics (NSI) [21]—average annual fruit consumption expressed in kg per capita (FRUIT_CONS), average annual fruit consumption expressed in daily calories per capita (FRUIT_CONS_CAL), average fruit production expressed in kg per capita (FRUIT_PROD), average meat and meat products consumption in the fresh meat equivalent expressed in kg per capita (MEAT_CONS), monthly average inflation rate for food goods expressed in % (INFLATION_FOOD);

(b).

FAO [66]—fruits export quantity, fruits excluding wine, expressed in 1000 tonnes (FRUIT_EXPORT), annual temperature change expressed in °C for the meteorological year (TEMPERATURE_CHANGE);

(c).

World Bank [67]—GDP per capita expressed in constant 2010 USD (GDP), total GDP expressed in constant 2010 USD million (GDP2), net inflows of foreign direct investment expressed in % of GDP (FDI), employment in agriculture expressed in % of total employment (EMPLOYMENT), government expenditure on education expressed in % of GDP (GOV_EXP_EDU), real effective exchange rate index expressed in 2010 = 100 (EXCHANGE_INDEX); according to the World Bank, “EXCHANGE_INDEX” is the nominal effective exchange rate (a measure of the value of a currency against a weighted average of several foreign currencies) divided by a price deflator or index of costs” [68].

(d).

European Commission [69]—emissions of greenhouse gases in CO2 equivalent (CO2, N2O, CH4, HFC, PFC, SF6, and NF3) from agriculture expressed in thousand tonnes (GHG_AGRI).

Given the fact that some variables were not found in previous studies, we considered adding them to the studies, as we expect them to have an influence on fruit supply (annual temperature change, foreign direct investments, and employment in agriculture).

The data were tested for stationarity [70], normality and autocorrelation [71,72,73] before exploring the correlation matrix and the time series models. Thus, some variables were transformed through first difference in order to ensure the compliance with these three criteria for the data mentioned previously. The transformed variables received in their coded names “D_”, which indicates the change of the considered variables from the previous year. Further, the correlation matrix was created for indicating the existence of relationships between the variables. However, this does not prove which impact each independent variable had on the dependent variables (D_FRUIT_CONS or FRUIT_CONS_CAL). Therefore, the Granger causality test and the regression models of time series were tested for this purpose. The Granger test indicates which independent variables impact the dependent variables by providing insight into the direction of the causality or, more concretely, the response or feedback from some kind of shocks from one variable to another, while the regression model specifies the dimension of the impact of independent variables on the dependent one [71,72,73].

The general equation of the relevant regression models found through testing are described in Equation (1),

$$dependent\_variable{s}_t={\displaystyle \sum }_{t=j}^n(a_t\times independent\_variable{s}_t)+{\epsilon }_t$$

(1)

where:

$dependent\_variable{s}_t$ = the change of the average annual fruit consumption expressed in kg per capita (D_FRUIT_CONS) OR the average annual fruit consumption expressed in daily calories per capita (FRUIT_CONS_CAL).

$a_t$ = the constant value attributed to each independent variable considered in the model.

$independent\_variable{s}_t$ = different combinations of the following: average fruit production expressed in kg per capita (FRUIT_PROD), average meat and meat products consumption in the fresh meat equivalent expressed in kg per capita (MEAT_CONS), monthly average inflation rate for food goods expressed in % (INFLATION_FOOD), Fruits export quantity, fruits excluding wine, expressed in 1000 tonnes (FRUIT_EXPORT), annual temperature change expressed in °C for the meteorological year (TEMPERATURE_CHANGE), the change of GDP per capita expressed in constant 2010 USD (D_GDP), the change of total GDP expressed in constant 2010 USD million (D_GDP2), the change of net inflows of foreign direct investment expressed in % of GDP (D_FDI), the change of employment in agriculture expressed in % of total employment (D_EMPLOYMENT), the change of government expenditure on education expressed in % of GDP (D_GOV_EXP_EDU), the change of real effective exchange rate index expressed in 2010 = 100 (D_EXCHANGE_INDEX), and the change of emissions of greenhouse gases in CO2 equivalent from agriculture expressed in thousand tonnes (D_GHG_AGRI).

${\epsilon }_t$ = the error attributed to each model; t = time frame from the year j to the year n.

4. Results and Discussions

4.1. Trends in Fruits Relevant Studies and Indicators

4.1.1. Results of the Bibliometric Analysis

From the 1040 articles obtained from the Scopus database, dating since 1980, the greatest attention paid to fruit consumption was registered in 2020, with 19.33% of the total results. Meanwhile, in the Web of Science (WoS) publications, the first articles appeared in 1991, and then the interest on the topic of fruit consumption started to increase. The highest number of publications from WoS was registered in 2019, with 59 publications, respectively, which is 8.35% of the total 707 results obtained. In addition, out of the total results of 1040 works from the Scopus database, 40.67% were open access and over 93% were written in English. At the same time, 96.25% of the results were published in a journal, and 87.88% were research articles. While, of the total publications obtained through the Web of Science platform, 48.23% were open access, 88.68% represent research articles and over 96% were written in English.

The most frequent fields of interest of the Scopus results are medicine (71.15%), nursing (25.58%), agricultural and biological sciences (19.42%), biochemistry, genetics, and molecular biology (19.04%), and environmental science (6.35%). The country with the most publications on Scopus is the United States of America (253 publications); Romanian-related results registered only eight publications. From the Web of Science results, 75.53% are found in the top five research areas: nutrition dietetics (25.18%s), public environmental occupational health (23.9%), oncology (13.01%), agriculture (7.5%), and food science technology (5.94%). Most of the WoS results come from the United States of America, 26.03% of them, while Romanian-related results record only three publications.

The authors with the most publications on Scopus are: La Vecchia, C. (14 publications), Boeing, H. (13 publications), Brug, J. (13 publications), Cheie, TJ (13 publications), Pengpid, S (13 publications). The authors with the most publications on WoS are: De Vries H (12 publications), Brug J (11 publications), Ball K and Tsugane S (10 publications each). La Vecchia’s studies [74,75,76,77,78] concentrate on the consumption of fruits from the perspective of human health benefits, such as the association of fruit and vegetables consumption with alleviating various diseases or the relationship between fruit and vegetables consumption and the intake of nutrients. Boeing focused his work [14,79,80,81,82] on demonstrating that a balanced diet that also contains fruits and vegetables helps the body in preventing health problems and diseases. Brug [83,84,85] emphasizes the need for an adequate consumption of fruits and vegetables for children and adolescents, and the ways to promote consumption among schoolchildren, along with lifestyle determinants of consumption preferences, especially in the consumption of fruits and vegetables. Pengpid [86,87,88] studied the determinants in the occurrence of chronic diseases and the influence of fruit consumption in reducing risks to human health.

The first highest cited articles on Scopus, according to the previously mentioned search criteria, from the fields of agricultural economics policy, business, management, economics fields are discussed in the following part.

Trienekens et al. [89] created a high-performance and innovation-based supply chain model based on the decisions of agri-food companies. Sijtsema et al. [90] surveyed the relationship between fruit consumption and its influencing factors, emphasizing a greater attraction towards eating fruits by people who prefer a sourer taste of food. In addition, several authors have investigated the problem of low fruit consumption and determined a series of influencing factors for this issue, such as socio-cultural values and factors [91,92,93] and economic ones [94]. For example, Goryakin et al. [94] showed that the low financial resources and the low availability of these products in nearby stores are main factors influencing low fruit consumption. Finally, it seems that an increase in fruit consumption can improve both the quality of the diet and the sustainability of the environment [95], and that innovation has an important role in this direction [96].

Figure 1 illustrates the keywords map of the investigated literature from the Scopus database.

Figure 1 shows the links between the 869 keywords selected as having at least five occurrences in the corpus. In the top of the most common keywords we found: “human” with 811 appearances, “female” with 686 appearances, “article” with 662 appearances, “male” with 630 appearances, “humans” with 559 appearances, and only in sixth place is “fruit”, with 570 appearances. In this context, it may be considered that the social aspect of sustainability is more debated when discussing topics related to fruit consumption.

Figure 2 illustrates the keywords map of the investigated literature from Web of Science database.

Figure 2 reflects the links between the 243 keywords generated by following the previously mentioned occurrence criterion and it highlights the words according to the highest number of their occurrences. The top of the most frequent keywords consists of: “fruit” with 104 appearances, “diet” with 100 appearances, “physical-activity” with 86 appearances, and the 5th and 6th are the words “risk” and “risk- factors” with 85 appearances and 84 appearances, respectively. As well as in the case of Scopus publications, the WoS results focus on the social aspect of sustainability when discussing fruit consumption related topics.

4.1.2. Descriptive Statistics of Fruit-Relevant Indicators Used in the Tested Models for Romania within EU Context

Regarding the consumption of fruits/capita in Romania reported by Eurostat [69], there is a significant increase registered in 2017 compared to 1995, an increase of approximately 83%, as illustrated in Figure 3.

The entire analyzed period was overall ascending, the maximum value being registered in 2016 with 90 kg of fruits/capita while in 1990 the average consumption of fruits/capita was only 48.31 kg. Regarding Romania’s position at EU-28 level, it ranks 23rd (63.92 kg/capita) in terms of the average amount consumed annually during 1995–2017, the top of the ranking being occupied by countries such as Luxembourg (165.89 kg/capita), Greece (142.53 kg/capita) and Italy (138.75 kg/capita). Carrying out an analysis at the level of 2017, the consumption in Romania increases, occupying the 10th place (88.25 kg/capita) in a ranking led by Portugal, Greece, and Slovenia (countries whose fruit consumption/capita exceeds 118 kg) [12]. The fruit production/capita in Romania registers oscillations in the period 1995–2017. By reporting the year 2017 to 1995, we notice an increase of only 5%, while the maximum value of production/capita was recorded in 2003, reaching 146.43 kg/capita. Considering the ranking of fruit production/capita at the EU-28 level for the period 1995–2017, Romania is in the 10th place, with an annual average of approximately 108.55 kg/capita, at the top being Spain (364.31 kg/capita), followed by Greece (335.82 kg/capita) and Cyprus (298.76 kg/capita). Compared to 2017, Romania occupies the 8th place (104.41 kg/capita), while Spain maintains its leading position with 306.32 kg/capita. The oscillation of fruit production in Romania from one year to another is caused by the main factors with significant impact on agriculture, that were further considered in the analysis, such as: weather conditions [97], the low interest of investments [98], and the change in agricultural employment [99].

In addition, organic farming is considered to be an important component in the sustainable development of the agricultural sector. Analyzing the data provided by Ministry of Agriculture and Rural Development (MARD) [100], it seems that the organic area with fruit trees had an upward trend between 2010 and 2018. Furthermore, it recorded an increase of 500% in 2018 compared to 2010, while the total area in organic farming had an increase of 78.57% in 2018 compared to 2010. In the case of organic farming, the indicator for determining the structure of agricultural land by categories of use had a superior value compared to the results in conventional system, as the percentage of ecological area in total agricultural land was 5.69% in 2018 compared to the percentage of conventional cultivated area in total agricultural land, which was 2.49%.

Moreover, based on the evolution of the agricultural area and of the area destined for fruit production in Romania in the period 1990–2018, an indicator was calculated to determine the structure of agricultural land by categories of use as a percentage ratio between the area for fruit production and the total agricultural area. The use of land with fruit trees could represent an indication of sustainable approach, if the land is used extensively, rather than intensively. The evolution of the indicator may be seen in Figure 4.

In the period 1990–2006, the agricultural surface of Romania was in the range of 14–15 million hectares, following a decrease in the next 3 years, reaching in 2010 an area of 14,156,000 hectares, but following that, in the period 2010–2018, its value remained below 14 million hectares. In 1990, Romania had 488,000 hectares for fruit production. Starting with 1991 and until 1999, there was an increase that kept over 500,000 hectares in production, but from 2000, that decreased to 334,000 hectares in 2018. If the agricultural area decreased by 9.17% in 2018 compared to 1990, in terms of areas cultivated for fruit production, there is a decrease of 31.57%. Considering the low value of the indicator calculated over the entire time interval analyzed to identify the type of land use, there is an extensive use of cultivated areas for fruit production in Romania. From the point of view of sustainability, the extensive use of land means a more sustainable approach of using the land with fruit trees, as in the case of Romania, than the intensive use, which is not the case. The indicator had an upward trend in the period 1990–1993, following small fluctuations until 1997, when it began to decline. Therefore, we may observe that instead of supporting the local production of fruits, Romania had a significant decrease in fruit production. In this case, the food policies should be aimed at supporting local fruit production so as to indirectly favor the increase in fruit consumption and contribute to a more sustainable development.

Further,

Table 1. Descriptive statistics for the Romanian indicators aimed to be considered in the models.

Indicators	Mean	Maximum	Minimum	Years	Time Frame
FRUIT_CONS (kg/capita)	66.89	111.30	43.40	30	1990–2019
FRUIT_CONS_CAL (daily calories/capita)	127.47	217.00	80.00	30	1990–2019
FRUIT_PROD (kg/capita)	63.58	96.80	40.40	30	1990–2019
FRUIT_EXPORT (1000 tonnes)	98.59	180.00	27.00	29	1990–2018
MEAT_CONS (kg/capita)	56.45	74.40	42.90	30	1990–2019
GDP (constant 2010 USD) (per capita)	7117.81	12,079.55	4352.13	30	1990–2019
GDP2 (constant 2010 USD million)	148,858.20	234,000.70	99,203.72	30	1990–2019
FDI (% of GDP)	2.96	9.02	0.00	30	1990–2019
INFLATION_FOOD (%)	2.50	12.00	−0.50	27	1991–2019 (no 2009 and 2014)
EXCHANGE_INDEX (2010 = 100)	85.77	112.66	42.10	29	1991–2019
EMPLOYMENT (% of total employment)	32.87	45.21	21.24	29	1991–2019
GOV_EXP_EDU (% of GDP)	3.40	4.31	2.87	19	2000–2017 + 1996
TEMPERATURE_CHANGE (°C)	0.99	2.28	−0.42	30	1990–2019
GHG_AGRI (thousand tonnes)	20,888.75	33,917.01	17,321.33	30	1990–2019

Source: authors’ own interpretation with EViews.

synthetizes the descriptive statistics for the Romanian indicators considered in the regression models and gathered from the databases previously presented in the Materials and Methods section.

During the analyzed period, the consumption of fruits in Romania registered an average value of 66.89 kg/capita (127.47 daily calories/capita), the maximum value was reached in 2019, 111.3 kg/capita (217 daily calories/capita), while the minimum value was recorded in 1999, a value by about 35 percentage points less than the average value of kilograms of fruit consumed/capita and by 37 percentage points less than the average value of calories from fruit in a day/capita. The year 1999, the year in which the lowest fruit consumption/capita was registered, also registered the lowest quantity of fruits exported from the analyzed post-communist period; these two indicators being directly influenced by the small fruit production/inhabitant registered in that year, a production of only 41.7 kg/capita. The highest fruit production was registered in 2003, with 52% more per capita than the average value registered in the analyzed period, while the year 1995 registered the lowest fruit production per capita, with 36% less than the average value of the analyzed period.

The value of GDP in relation to the number of inhabitants registered the highest value in 2019, in terms of meat quantity consumed and fruit quantity consumed per capita, and the minimum value of GDP/capita was registered in 1992, the year in which the consumption of fruits and meat per capita registered significantly low values, below the average of the analyzed period. Based on these links between indicators, we could say that there is a directly proportional relationship between GDP/capita indicators and fruit and meat consumption.

4.2. Determinants of Fruits Consumption in Romania—A Time Series Approach

In order to provide an overview on the sustainability of consumption in the fruit sector, an understanding of the macroeconomic determinants of fruit consumption is needed. Then, a discussion on the sustainability of consumption could be generated. The correlation matrix of the considered variables may be seen in Table S1 in the Supplementary Materials.

The change of fruit consumption expressed in kg per capita (D_FRUIT_CONS) is in relation with the following variables: the daily calories intake per capita (FRUITCONS_CAL), the average annual fruit production per capita (FRUIT_PROD), the change of net inflows of foreign direct investment (D_FDI), the change of meat consumption per capita (D_MEAT_CONS), the change of daily calories intake of fruits per capita (D_FR_CONS_CAL), and the change of total GDP (D_GDP2), as the probability is significant under 0.10. All of them registered a strong positive relationship with the considered dependent variable—D_FRUIT_CONS, as indicated by the coefficient of the relationships, which are higher than 0.30. For the rest of the independent variables, it seems that the correlation coefficients are not significant and further testing should be performed in future studies on different and bigger samples.

In what concerns the fruit consumption expressed in daily calories per capita (FRUIT_CONS_CAL), a relationship with the following variables was found: the change of fruit consumption expressed in kg per capita (D_FRUIT_CONS), the average annual fruit production per capita (FRUIT_PROD), meat consumption per capita (MEAT_CONS), the change of GDP per capita (D_GDP), food inflation (INFLATION_FOOD), the annual temperature change (TEMPERATURE_CHANGE), GHG emissions from agriculture (GHG_AGRI), the change of meat consumption per capita (D_MEAT_CONS), the change of daily calories intake of fruits per capita (D_FR_CONS_CAL), and the change of total GDP (D_GDP2), as the probability is significant under 0.10. Concerning most of the variables, a positive relationship was registered, except with INFLATION_FOOD, and the GHG emissions from agriculture (GHG_AGRI). For the rest of the independent variables, it seems that in this case the correlation coefficients are not significant and further testing should be performed in future studies on different and bigger samples.

Further,

Table 2. The Granger test for the significant causality between indicators.

Null Hypothesis:	Obs.	F-Statistic	Prob.
D_FRUIT_CONS does not Granger Cause D_FR_CONS_CAL	27	3.14008	0.0631
D_FRUIT_CONS does not Granger Cause D_GDP	27	3.03041	0.0688
D_FRUIT_CONS does not Granger Cause D_GDP2	27	3.23323	0.0587
D_FRUIT_CONS does not Granger Cause INFLATION_FOOD	21	3.92925	0.0409
D_FRUIT_CONS does not Granger Cause D_EXCHANGE_INDEX	26	2.58838	0.0989
D_EMPLOYMENT does not Granger Cause D_FRUIT_CONS	26	5.34786	0.0133
D_EMPLOYMENT does not Granger Cause FRUIT_CONS_CAL	26	5.55194	0.0116
FRUIT_CONS_CAL does not Granger Cause TEMPERATURE_CHANGE	28	4.96887	0.0161

Obs. = observations; Prob. = probability; Source: authors’ own interpretation with EViews.

presents only the results for which the Granger test is significant in providing proof of some kind of feedback when considering the two considered dependent variables—D_ FRUIT_CONS and FRUIT_CONS_CAL.

The Granger causality test is significant only in the cases described in Table 2. It seems that the change of fruit consumption expressed in kg per capita (D_FRUIT_CONS) causes some effects in the food inflation rate, GDP total and per capita, and the modification of the exchange index from one year to the next one. At the same time, fruit consumption expressed in daily calories per capita (FRUIT_CONS) causes some effects in the temperature change. Furthermore, the change in the employment from agriculture from one year to the next seems to have some effects on both the change of fruit consumption expressed in kg per capita (D_FRUIT_CONS) and the fruit consumption expressed in daily calories per capita (FRUIT_CONS_CAL). Although the other independent variables seem not to Granger cause the dependent ones, further investigation should be conducted, as [71] showed.

Onward,

Table 3. Estimated models for the increase in fruit consumption expressed in kg/capita (D_FRUIT_CONS) and fruit consumption expressed in average daily calories (FRUIT_CONS_CAL).

Variables	Fruit Consumption Change from One Year to the Next (kg/capita) D_FRUIT_CONS			Fruit Consumption (Average Daily Calories) FRUIT_CONS_CAL
	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6	Model 7
	1992–2019	1992–2019	1996–2017	1990–2019	1990–2019	2001–2017	1990–2019
FRUIT_PROD	0.337 *			1.432 *
MEAT_CONS					3.626 *
D_GDP			0.0003 ***			0.045 si (p = 0.14)
D_FDI	1.305 **
D_EXCHANGE_INDEX	−0.203 ***
D_EMPLOYMENT		−1.291 ***
D_GOV_EXP_EDU						5.888 si
TEMPERATURE_CHANGE	−1.197 si	−3.516 ***		29.771 *
GHG_AGRI							−0.004 **
D_INFLATION_FOOD						30.067 **
D_MEAT_CONS	0.604 **
c—constant	−18.480 *	5.531 **	0.444 si	6.982 si	−77.229 *	121.15 *	216.81 *
R-squared/adjusted	0.73/0.67	0.19/0.13	0.10/0.06	0.60/0.57	0.79/0.78	0.54/0.39	0.15/0.12
F-statistic	12.05 *	3.11 **	3.01 ***	20.26 *	106.09 *	3.65 ***	5.21 **
Durbin-Watson	2.004	2.15	2.18	1.16	0.68	1.47	0.29

*, ** and *** significance at <1%, 1–5%, and >5–10%, ^si—statistically insignificant. Source: Authors’ own interpretation with EViews.

presents the results of some regressions models for the two dependent variables though testing the combination of independent variables.

The fruit consumption, which increases from one year to the next, and is expressed in kg/capita (D_FRUIT_CONS), is negatively impacted by the following variables: the increase in the exchange rate index, the increase in the employment in agriculture, and the temperature increase. Specifically, the increase with 1 RON/EUR (2010 = 100) of the change from one year to the next one of the real effective exchange rate index would generate a decrease in fruit consumption of 0.203 kg/capita, which supports previous results [59], considering that the purchasing power must be supported through national policies against external exchange rate fluctuations. The increase of 1% of the change from one year to the next of employment in agriculture as a percentage in total employment would generate a decrease in the fruit consumption of 1.291 kg/capita. While the employment in agriculture is a direct influence of the fruit supply [99], the increase in employment is natural to generate consumption decreases, as the labor would try to suffice the lack of technology and innovation in agriculture, therefore having a negative impact. The increase of 1 °C of temperature would generate a decrease from one year to the next for fruit consumption of 3.516 kg/capita, the results being in line with previous studies [63]. This means that fruit consumption would have a decrease in time, from one year to the next, if the temperature change were to increase. This might indicate a greater attention of consumers to climate change topics and, perhaps, a better understanding of their impact from a consumption behavior viewpoint. However, this should be the focus of a future study in order to investigate if this assumption is correct or it is only a contextual pattern generated by other factors.

Further, the daily intake of calories from fruit consumption (FRUIT_CONS_CAL) increases, influenced by the following variables: increase in fruit production, meat consumption, the increase in GDP per capita (if we accept an error of approximately 14%, as p = 0.14), the temperature increase, and the increase in food inflation, as their impact coefficients are positive. While a decrease in fruit consumption may be generated by the following variables: the increase in the exchange index, the increase in the employment in agriculture, and temperature increase, as their impact coefficients are negative.

Specifically, the yearly increase of 1 kg per capita of fruit production would generate an increase of 1.432 daily calories/capita from fruit consumption. The yearly increase of 1 kg per capita of meat consumption would generate an increase of 3.626 daily calories/capita from fruit consumption; again, these results are in line with previous studies [31,62]. This might be explained by the higher affordability of food. Surprisingly, the increase of 1 °C of temperature change would generate an increase of 29.771 daily calories/capita from fruit consumption. However, the yearly increase in GHG emissions from agriculture of 1 thousand tonnes per capita would generate a decrease of 0.004 daily calories/capita from fruit consumption, the second part of this result being in line with Heller et al. [63]. These two previous results indicate a need for a better investigation of environmental factors’ impact on consumption. Finally, the increase in the yearly change of food inflation of 1% would generate an increase of 30.067 daily calories/capita from fruit consumption, supporting previous results [59].

Moreover, the fruit export quantity does not explain the two dependent variables—D_FRUIT_CONS and FRUIT_CONS_CAL—which is why, although tested, it was not added to the results. In terms of trade, maybe other indicators could influence fruit consumption, such as trade as percentage of GPD, but this might be a new subject for a future study. Alternatively, other models would be more suited for this relationship than the linear regression, which was tested in this paper. Similarly, food inflation could not be represented in the analyzed regressions, as it was strongly correlated with other independent variables (see the matrix of correlation), which would have infringed the hypothesis that independent variables, which explain the dependent variable, should not be correlated with each other, because it would generate invalid models [72,73].

When the R-squared and the adjusted R-squared goes towards 1, such as in model 1 and model 5, a strong influence appears of the independent variables on the two dependent ones: D_FRUIT_CONS and FRUIT_CONS_CAL. Comparatively, the fourth and sixth models registered a moderate impact. In addition, in the second, third and seventh models, the R-squared indicates a weak impact of the independent variables considered in the models onto the dependent variables. Furthermore, the statistical significance of the F-statistic test, represented by *, **, and ***, shows that all models are valid. In addition, the models which have the value of the Durbin-Watson test near two are the most accurate; as there is no residual autocorrelation, an assumption needed to be met for regression validity [72,73].

However, other factors should be tested in the future in order to see how they contribute to the evolution of the dependent variables for a better understanding of fruit consumption’s sustainability.

5. Conclusions

The aim of the paper was to analyze some macroeconomic factors with the potential to influence fruit consumption in Romania. The main results obtained indicate an increased interest in research on sustainability as well as fruit consumption and its determinants, the results being obtained through the application of bibliometric analysis.

The statistical analysis highlights the upward trend of fruit consumption per capita in Romania, but at the EU-28 level, Romania ranks among the last places in terms of fruit consumption per capita. The results indicate that from the selected macroeconomic factors, fruit consumption is positively influenced by the increase in fruit production, the increase in GDP/capita, the increase in foreign investments as % of the GDP and the increase in meat consumption. The highest increase is determined by the increase of 1% in foreign investments as percentage of the GDP. Furthermore, a negative impact on fruit consumption (a decrease) is generated by the increase in the exchange rate, increases in the employment in agriculture as percentage of total employment and temperature increases. The highest negative impact being generated by the increase in agricultural employment.

Given the results of this analysis, there is a need for national authorities to put together a set of policies and measures to reach the objectives of Agenda 2030 [17]. By having an overview of the macroeconomic factors that influence fruit consumption, some measures aimed at increasing fruit consumption may be generated, for example financial support for local fruit production so that production increases, along with employment decreases and technology increases, which would generate a higher supply and therefore a higher consumption. Other measures should focus on support for reducing the greenhouse gas emissions and reducing the speed of temperature increase, through financial support or knowledge exchange, if possible. Having control on the inflation rate, food prices and attracting foreign investments should also be considered as necessary measures for supporting the consumption of fruits and implicitly a sustainable development at national level. What needs to be mentioned is that the correlations and influences between variables have significant importance in developing future measures and policies, since they reveal indirect levers for reaching common development goals.

The limits of the research are the low availability of data specific to the fruit sector to provide information on sustainable fruit consumption, data that would have helped to develop the descriptive analysis. Furthermore, the data series is restricted to the selected period because statistical data regarding the agricultural production in Romania before 1990 is scarce and dissipated. For future studies, the consumption of ecologically certified fruits from Romania is pursued as part of sustainable consumption. In addition, further studies could investigate other determinants of agricultural and food products’ consumption, such as meat, as its contribution to climate change is acknowledged in the literature, and we kindly invite other scholars to contribute to developing this area.