The Impact of Trade and Consumption on Kosovo GDP Growth: An Analysis Based on the Cobb-Douglas Model

This chapter reviews the literature on trade, consumption, and GDP growth in Kosovo, analyzing the key factors influencing the country's economic development and its challenges in integrating into global markets. In the WIIW report for 2025, trade and consumption are key factors in GDP growth in Kosovo, where private consumption, driven by remittances, helps increase demand, but challenges such as high unemployment and dependence on imports negatively affect the trade balance and economic stability, while exports play an important role in economic growth, but difficulties related to underdeveloped infrastructure and lack of product diversification hinder this opportunity, while sectors such as agriculture and services offer further opportunities for trade development and economic growth [2]. Kosovo's economic progress can be analyzed through classical and neoclassical economic theories. Ricardo [11] argued that nations specialize in producing goods where they have a productivity advantage, leading to increased consumption and enhanced global welfare. While neoclassical theory emphasizes the effect of trade on wage inequality across sectors, newer theories focusing on strong heterogeneity examine its impact on wage distribution within sectors and occupations [31]. Heckscher and Ohlin [32] also analyzed trade, suggesting that countries tend to export goods that rely on abundant domestic factors of production, while importing goods that require relatively scarce resources. He emphasized the role of abundant factors, arguing that countries with more resources would export goods that require intensive use of these factors [33]. Krugman’s [34] new trade theory offers two main insights: first, trade depends on the similarity of countries' factor endowments, with countries becoming more similar engaging in intra-industry trade. Second, when intra-industry trade dominates, the benefits of market expansion will outweigh distributional effects, benefiting both scarce and abundant factor owners. Firms with different productivity levels coexist in an industry due to initial uncertainty about their productivity before making an irreversible investment. Entering the export market is costly, but the decision to export is made after firms gain knowledge about their productivity [35]. Empirical studies, such as those by Sachs and Warner [36], confirm a positive relationship between trade openness and economic growth, which includes increased productivity and the creation of new employment and consumption opportunities. Trade openness can stimulate economic growth by allowing countries to exploit comparative advantages and increase productivity through specialization [37]. Increased competition from trade can also drive innovation and improvements in government policies and institutions [38]. Firms can benefit from market expansion beyond national borders, leveraging economies of scale and increasing investment in research and development. For instance, the East Asian Tigers utilized exports to achieve rapid capital growth and improvements in capital-intensive industries during the 1970s and 1980s [39]. Empirical evidence supports the net benefits of trade, including productivity growth, innovation, and technological improvements through specialization and comparative advantages [40]. The data analysis shows that during the period from 1950 to 1998, countries that liberalized their trade regimes experienced average annual growth rates about 1.5 percentage points higher than before liberalization. Investment levels increased by 1.5-2.0 percentage points after liberalization, confirming that liberalization stimulates growth partly through its effect on physical capital accumulation. Additionally, liberalization raised the trade-to-GDP ratio by about 5 percentage points, indicating that trade policy liberalization effectively increased the level of openness in these countries [41]. Trade is a key driver of productivity and growth, and its impact depends on its relative importance in the overall economic activity [42]. However, some critics suggest that other factors may also influence growth [28], including consumption as a key factor. Based on classical and neoclassical trade theories and the Cobb-Douglas model, we can analyze the impact of imports, exports, and final consumption on GDP growth. In this context, we aim to determine whether the growth of exports has a significant impact on Kosovo's economic growth, while also examining the role of domestic consumption and imports in capital accumulation. However, it is important to also consider other factors, such as Kosovo's economic structure and its development policies, which could play a crucial role in future growth. Kosovo's economy, while showing progress in macroeconomic stability and GDP growth, faces challenges such as high unemployment 17.2% in 2023, trade imbalances, and dependence on remittances around 17% of GDP in 2023 [43], which require a strategic approach grounded in classical and neoclassical economic theories and the Cobb-Douglas model to address structural issues and unlock growth opportunities. Kosovo should focus on developing export-oriented sectors like agriculture, IT, and renewable energy, but challenges such as low labor productivity €10,542 per worker in 2023, and since 1 January 2024, holders of Kosovo passports enjoy visa-free travel to the EU [6] and limited industrial development (12% of GDP) hinder its competitiveness [44]. According to data from the Central Bank of Kosovo (2007-2024, September), around 7 billion euros have been brought to Kosovo. The countries that have invested the most in Kosovo are Germany (19.8%), Switzerland (9.9%), the United States (7.3%), Albania (6.6%), Austria (5.2%), the Netherlands (4.7%), Turkey (3.9%), France (3.7%), and Slovenia (3.6%), while other countries contribute 47.2% of the total foreign direct investment in Kosovo. Only in 2024, investments have been brought to Kosovo with the aim of improving infrastructure to ensure sustainable growth [45]. The Solow-Swan neoclassical growth model highlights that capital accumulation, labor force growth, and technological advancements are key drivers of GDP expansion, which is reflected in Kosovo's 4.5% GDP growth in early 2024 [46]. This growth, supported by remittances, consumption, investments, and credit expansion, shows the importance of these factors in fostering economic development. However, challenges such as limited production diversification and a 2.75% current account deficit in 2024 emphasize the need to improve domestic production and reduce reliance on imports [47-49]. The Cobb-Douglas Model (Factor Contribution) represents economic output as a function of capital (K) and labor (L), explaining their impact on GDP growth. The rise in private sector lending, which grew at 13.2% in 2023 [47], along with investments and consumption, demonstrates how capital and labor contribute to economic expansion. However, high unemployment, informality (which accounts for approximately 30% of the economy [50]), and slow productivity growth limit efficiency and prevent the economy from reaching its full potential. To improve Total Factor Productivity (TFP), Kosovo should focus on technological advancements, workforce skill development, and investment in high-value sectors. Economic sustainability can be achieved through well-structured economic policies that integrate comparative advantage, capital accumulation, and production efficiency. By addressing these structural issues and focusing on key sectors, Kosovo can achieve long-term growth and stability. Kosovo can benefit from classical and neoclassical economic theories to improve its development. According to Ricardo [11], countries should specialize in sectors where they have a comparative advantage, focusing on areas such as agriculture, information technology, and renewable energy [51, 52]. Ruffin [53] emphasizes the importance of integrating into global trade and utilizing the labor force and natural resources, while George Stigler connects competition and market efficiency to sustainable economic development [54]. Kuznets [55] suggests that development requires structural changes, shifting the economy from consumption and imports to production and exports [56]. Solow [56] highlights the role of capital, labor force, and technology for sustainable growth, and Romer [57] expands this by emphasizing the importance of innovation and knowledge. Stiglitz [10] stresses strengthening institutions and reducing inequalities for equitable development, while Keynes [58] advocates for government intervention to manage demand and avoid instability. Finally, Krugman [59] links economic development to scale and specialization in sectors where competitive advantages can be achieved. This combination of theories is essential for guiding Kosovo's economic development. For Kosovo, the most appropriate theories are those that emphasize comparative advantage, the importance of developing strategic sectors, and the need for structural reforms to transition from a consumption and import-based economy to one focused on production and exports. David Ricardo's Theory of Comparative Advantage is particularly relevant for Kosovo, as it offers a framework for identifying sectors where the country could have a competitive advantage, such as agriculture, information technology, and renewable energy. This focus could help boost exports and strengthen trade ties with other countries. Furthermore, the Solow-Swan neoclassical growth model and Romer's growth theory are also important, as they highlight the need for investment in capital, infrastructure, and technology to increase productivity and foster long-term growth. In this context, combining these theories could be beneficial for Kosovo, focusing efforts on increasing competition, diversifying production, and targeting sectors that can generate comparative advantages and promote export growth.

2.1 The impact of exports on GDP

The Exports are widely regarded as a key driver of economic growth, boosting production, creating jobs, and increasing national income. In Kosovo, however, exports have remained modest, contributing minimally to GDP growth. The country's mining, energy, and agriculture sectors, which could provide comparative advantages, remain underdeveloped, hindering its potential as an export-driven economy. A narrow export base, dominated by low-value-added goods, reflects Kosovo’s failure to fully utilize Smith’s concept of specialization and competitive advantage. Neoclassical trade theory highlights that exports increase demand for a country's products in international markets, driving GDP growth. Balassa [60] argued that exports optimize resource allocation, utilize economies of scale, and stimulate technological advancements in response to global competition. These processes also help address surplus labor and support employment growth. Feder [61] analyzed semi-industrialized countries from 1964 to 1973, showing that productivity is significantly higher in export sectors due to cross-sectoral externalities. He concluded that reallocating resources from less efficient sectors to export-oriented ones could enhance growth beyond traditional inputs like labor and capital. Trabelsi and Kachout [62] found that agricultural exports have a negative and insignificant impact on economic growth in both the short and long term. Other exports negatively and significantly impact growth, while imports positively influence long-term economic growth. These findings suggest the need to focus on promoting specific sectors to drive growth. Bakari et al. [63] analyzed Peru's economy and concluded that domestic investments, exports, and imports do not significantly affect economic growth in the short or long term. They attribute this to poor economic organization, indicating that trade openness and domestic investment are not primary growth drivers in the Peruvian context. In contrast, Bakari et al. [64] examined Brazil’s economy from 1970 to 2017 using VECM methodology. They found that in the short run, imports, exports, and domestic investment positively contribute to economic growth, demonstrating their causal role in Brazil's economic expansion. According to Ninka and Pere [65], the gravity model of exports for the Western Balkan region shows that exports are positively influenced by shared language, common borders with third countries, trade with the European Union, and trade with large, highly industrialized countries. However, factors like distance and the region's per capita GDP have a negative impact on exports.

These studies collectively highlight the varied effects of trade, investment, and exports across different economies, emphasizing the importance of tailored economic strategies for growth.

2.2 The impact of imports on GDP

Imports can have both positive and negative effects on GDP. While they meet consumption and production needs and can enhance technological efficiency, excessive imports leading to trade deficits can harm economic growth. Krugman [59] argued that East Asia's rapid growth between 1965-1990 was driven by innovation, contrasting it with the stagnation of the Soviet Union, suggesting that Asia's growth is not a model for the West. He also noted that countries relying heavily on consumer goods imports face slower long-term growth. Kim et al. [66] found that imports positively affect productivity growth, especially through competition and technology transfers from developed nations. They argued that import liberalization can promote growth in certain circumstances. In Kosovo, Vardari [67] observed that exports were crucial for economic growth but their impact became insignificant when imports were included in the model, highlighting the significant role of imported intermediate goods in driving growth. Trade plays a crucial role in economic growth by fostering job creation, reducing poverty, and expanding economic opportunities, especially in developing nations [68]. This dynamic is reflected in the Albanian experience over the past decade, where a positive correlation between trade volume growth and GDP highlights the crucial role of trade in fostering economic growth and development [69]. Despite these benefits, trade flows in the region remain largely shaped by traditional economic ties rather than new initiatives like Open Balkan, which has yet to establish itself as a significant regional center of gravity [70]. Syzdykova et al. [71] analyzed Kazakhstan's economic data from 2000 to 2017, finding that both exports and imports contributed to GDP growth, although imports had a stronger short-term impact. Awokuse [72] reported similar findings in Eastern European countries, showing that both exports and imports positively influenced growth. In contrast, Trabelsi and Kachout [62] found that agricultural exports had a negative impact on growth, while other exports were negatively significant, with imports positively affecting long-term economic growth. These studies reveal a complex relationship between imports and GDP. Countries that import technological and capital goods can improve productivity and long-term growth. However, excessive imports without corresponding export capacity can lead to trade imbalances and hinder GDP growth, as seen in Kosovo's economy.

2.3 The impact of final consumption on GDP

When discussing the limitations of traditional economic indicators, Deaton [73] highlights growing concerns about the use of GDP as a measure of economic progress. Although there are no statistical issues with using GDP or GDP per capita, argues that the overemphasis on these measures by the media can undermine public confidence in national accounts. Instead, he advocates focusing on individuals' disposable income and consumption. This perspective is particularly relevant in Kosovo, where final consumption, including both government and private sector spending, plays a crucial role in GDP. In 2012, consumption was USD 6.19 billion and grew to USD 10.1 billion by 2023 [74], primarily driven by remittances from the diaspora [43]. However, despite this growth, poverty remains an issue, with GDP per capita at only USD 5 per day in 2023 and 19% of the population living in poverty in 2018 [74]. Furthermore, government spending makes up 12.5% of GDP, but the annual growth rate of Household Consumption Expenditure, which was 7.3% in 2021, slowed to 3.1% in 2023 [74], signaling a decline in the pace of household spending. While remittances stabilize consumption (USD 1.35 billion in 2024), they also create structural problems as they depend on the economic conditions of the host countries and do not reflect domestic productivity [75]. According to the neoclassical model, consumption should increase with higher productivity and income per capita. However, in Kosovo, remittances finance consumption without stimulating domestic production, limiting sustainable growth. Keynesian economics (1937) emphasizes that consumption drives aggregate demand, thus stimulating GDP in the short term. Increased disposable income leads to higher consumption, which stimulates domestic production. In Kosovo, where consumption is a large share of GDP, policies to increase purchasing power are crucial for growth. However, consumption growth exceeds domestic production, leading to increased imports and dependence on external factors, which hampers sustainable growth and weakens the trade balance. Milton Friedman's consumption theory suggests that individuals base consumption not only on current income but also on expectations for the future [76]. Folbre [77] explores the impact of time use on income and consumption inequality, examining empirical research on time distribution, leisure time, household production, and its implications for policy [77]. In Kosovo, remittances play a critical role in supporting consumption, but their uncertainty due to reliance on the economic conditions of host countries remains a challenge. To achieve more sustainable growth, Kosovo must develop competitive industries, increase exports, and reduce dependence on imports. Sudibyo [78] analyzed the contribution of consumption to economic growth in ASEAN3+ countries and the United States during 1960-2020, finding that consumption has a positive and significant effect on economic growth, while the Consumer Price Index (CPI) has a negative impact. Similarly, research on the impact of household consumption expenditures, government spending, and income per capita on economic growth in the Sidoarjo region, using secondary data for 15 years (2007-2021), shows that while household consumption expenditures and income per capita significantly affect economic development, government spending does not have a significant impact [79]. Studies also show that final consumption has a multiplier effect on GDP growth. Akermi et al. [23] demonstrated this effect, where increased consumption leads to further economic activity. Bakari et al. [64] found a positive relationship between consumption and growth in the U.S., while Perrotta [80] showed that this relationship is stronger in low- and middle-income countries. In a study by Almasifard and Saeedi [81], using panel regression, the authors found that the lagged value of final consumption expenditure as a percentage of GDP has a significant impact on the share of consumption in GDP in eight Central and Eastern European countries (1993-2010). Kosovo's focus on final consumption, primarily financed by remittances and imports, aligns with the theory that low-income economies allocate a large share of GDP to consumption. However, consumption theory criticizes such an economy for relying on unsustainable income sources that do not create a strong foundation for domestic investment in production. Kosovo's economy exhibits a mix of Keynesian elements through its dependence on consumption and possible government interventions, while also showing features of a liberal market economy, emphasizing trade liberalization and private sector development. Thus, the trajectory is not merely Keynesian or liberal but a hybrid model that responds to the needs of a developing economy. The country's high dependence on imports and remittances suggests reliance on external factors, but domestic policies aim to promote market liberalization for long-term growth.

This section presents empirical regression results and explores the impact and relationship of imports, exports and final consumption on annual GDP growth, using a time series data analysis for Kosovo, period 2007-2023. Descriptive statistical analysis presents the arithmetic mean and standard deviation (Table 2) of GDP annual growth, import, export and final consumption.

4.1 The descriptive statistic

The descriptive statistics are presented in Table 2 for the following indicators: GDP, FCE, EXP, and IMP (2007-2023).

Table 2. Descriptive statistics

	N	Minimum	Maximum	Mean	Std. Deviation
LagGDP	17	-11.17	10.8	2.1	4.4
LagFCE	17	71.36	106.6	83.6	8.5
LagIMP	17	23.93	104.7	54.0	21.1
LagEXP	17	17.00	99.3	47.1	24.6
N	17

Source: Data from the World Bank in US$

The analysis of the mean and standard deviation helps us understand the performance and variability of economic factors:

- LagGDP: The mean (2.15) indicates moderate economic growth, while the standard deviation (4.47) shows moderate fluctuations, suggesting stable growth but with periods of uncertainty.

- LagFCE: The mean (83.63) reflects steady and high spending, while the standard deviation (8.52) indicates moderate variability, possibly linked to factors like inflation.

- LagIMP: The mean (54.06) reflects average imports, and the standard deviation (21.13) suggests considerable fluctuations, likely influenced by international factors such as prices and trade policies.

- LagEXP: The mean (47.12) indicates average exports, and the standard deviation (24.65) shows high variability, reflecting changes in international demand and trade policies.

Overall, these statistics suggest an economy with moderate growth but significant variability in international trade.

4.1.1 The descriptive statistic

In recent years, Kosovo's economy has been impacted by external demand slowdown and rising inflation, but in 2024, economic activity grew by 4.5%, primarily driven by public investments, consumption, and service exports, while net exports had a negative effect on GDP growth, and inflation slowed to 1.6 % [111].

• Demographic analysis

Study variables that also shape the economic profile of Kosovo (https://bqk- kos.org/statistikat/serite-kohore/):

1. Nominal GDP: Nominal GDP in Kosovo Figure 1, in 2021 was 7,975 Million Euros in 2022 it was 8,936 Million Euros in 2023 it was 10,131 Million Euros and until March 2024 it was 2,095 Million Euros.

1.png

Figure 1. Quarterly percentage change in GDP, 2011-2024

Source: Central Bank of Kosovo (2024).

While the real GDP in 2022 was 7.499 million Euros or 6% lower than the nominal GDP and this trend continued in 2023 and Q1/2024. With Eq. (4) below, the changes in GDP for Kosovo are calculated. Even the real GDP growth rate in % (GDP Real growth rate) for 2021 was 10.7%, in 2022 it decreased to 4.3%, while in 2023 it was 3.23% measured by quarters. The year 2024, the period January-March, has an economic growth of 5.6% [44].

Elasticity or change in GDP for the country (j) is measured by Eq. (4):

$\Delta \mathrm{GDP}=\frac{G D P_t^J}{G D P_{t-1}^J}-1$            （4）

where, $\Delta$ GDP is a change in GDP between time $t$ and $t-1$, $G D P_t^J$ is GDP for country j at time ( t ) and $G D P_{t-1}^J$ is GDP for country jat time t-1.

The real GDP growth in Kosovo is an average of 4.5% for almost 2 decades [111].

• Kosovo: Main macroeconomic indicators

The inverted "V" shaped histogram shows that most values are concentrated at the extremes, with a low peak around the mean 4.54% (Figure 2). This suggests that there were years with both high and low GDP growth, while years of moderate growth were fewer. The data distribution shows significant variations, reflecting extreme economic events affecting GDP. The import of goods for 2014 was €2.5 billion, while a year later (2015) with €2.6 billion, in 2016 it was €2.8 billion, while in 2017 with €3.0 billion import.

2.png

Figure 2. GDP growth rate 2008-2024

Kosovo's economy (2008-2022) (Figure 3), Cluster Bar Mean of GDP shows GDP fluctuations between -5% and 10%, influenced by global crises, political uncertainty, and low investments. High imports (€51-71 billion) indicate dependence on foreign markets, while limited exports (€17-38.4 million) create a significant trade deficit. Structural improvements and increased production are essential for economic balance.

3.png

Figure 3. Cluster Bar Mean of GDP (2008-2022)

4.png

Figure 4. Trade balance

Source: Central Bank of Kosovo (2024) and KAS, (2024).

In Figure 4, slow import growth continues in 2018 with €3.3 billion, for 2019 it was €3.5 billion and for 2020 it was €3.2 billion. The import of goods for the year 2021 increased by €4.6 billion, while the year 2022 was €5.6 billion. The increase and decrease are also in 2023 with only €5.9 billion. The forecast of imports is expected to be around 6 billion euros, emphasizing Kosovo's dependence on imports. While the real GDP in 2022 was 7.499 million Euros or 6% lower than the nominal.

• Final Consumption

Final consumption expenditures in Kosovo have steadily increased from 2007 to 2023, reflecting higher household and government spending on goods and services, with significant growth between 2020 and 2023 due to post-pandemic recovery, government spending, and economic normalization after COVID-19 lockdowns. The overall trend shows moderate annual growth, particularly accelerating in recent years, driven by inflation, domestic spending, and recovery [111].

The Partial Regression Plot (Figure 5) analysis examines the relationship between GDP and final consumption (FCE) as dependent and independent variables.

From the regression equation:

y= −1.25×10−14−0.46⋅X

It shows a negative correlation between final consumption and GDP, where an increase in final consumption tends to lead to a decrease in GDP. However, this relationship is relatively weak, as R² = 0.144, indicating that only 14.4% of the variance in GDP can be explained by changes in final consumption. This result suggests that final consumption is not a dominant factor in explaining GDP movements and that other factors have a greater impact on the economy.

5.png

Figure 5. Partial Regression Plot for GDP and final consumption

6.png

Figure 6. Analysis Partial Regression Plot for PBB and imported

7.png

Figure 7. Analysis of the Partial Regression Plot for GDP and exports

8.png

Figure 8. The impact of exports, imports and FC on annual GDP growth in Kosovo

9.png

Figure 9. Histogram (Mean, Std.Dev., for impact of XM, IM and FC on GDP growth

The horizontal axis shows standardized residuals ranging from -3 to +3, while the vertical axis displays GDP values ranging from -15 to +15. The R² value is 0.716, indicating that 71.6% of the GDP variance is explained by the variables in the model. When the quadratic term is included, R² increases to 0.730, suggesting that the quadratic model is more appropriate. This shows that the relationship between various factors such as exports, imports, and final consumption with GDP is complex and better explained by a non-linear model. This analysis helps in formulating policies for the sustainable economic development of Kosovo.

The histogram (Figure 9) analyzes the relationship between imports, exports, consumption, and GDP. With a mean of 2.15 and a standard deviation of 4.465, the data shows significant variations from month to month. The vertical axis represents the frequencies, while the horizontal axis represents GDP. This graph helps in understanding the impact of economic factors on GDP, suggesting that there are substantial variations concerning the variables that influence Kosovo's GDP (Table 3).

Table 3. Olaps cube

Source: By authors

The OLAP cubes for FCE (Final Consumption Expenditures), EXP (Exports), IMP (Imports), and GDP offer insights into Kosovo's economic performance over 17 years. The GDP Sum is 74.30 %, representing total GDP growth, with a Mean of 4.53% and a Std. Deviation of 2.90%, showing moderate growth with considerable yearly fluctuations. The % of Total Sum for GDP is 100%, indicating all data points relate to GDP.

High mean values in EXP suggest export-driven growth, while high FCE values emphasize the importance of domestic consumption. Std. Deviation values highlight volatility, especially in trade. Analyzing these cubes helps understand Kosovo's economic dynamics and the contribution of consumption, trade, and GDP growth.

4.2 The statistic tests

This test is well known for its use in time series models to verify whether a change in one time series can predict a change in another series. Granger Causality test [104]. The Granger Causality test assesses whether the past values of one time series can predict the current values of another time series (Table 4).

Table 4. Granger causality results for FC, IM, and XM on GDP for Kosovo

This table presents the results of the Granger Causality test for the effect of consumption, exports, and imports on GDP, using a lag of 1 time period. The p-values from the test are all below 0.05, so we conclude that (X - imports, exports, and final consumption) Granger-causes an increase or decrease in (Y - GDP).

• Two-Sample KS Test

Table 5. Kolmogorov- Smirnov statistic test

The statistic (Table 5) indicates a relatively large discrepancy between the two sampling distributions. In our case, the p-value is below the chosen level of significance (usually 0.05), the null hypothesis is rejected. Both samples come from different distributions (p=value for each variable is, 2.08; 6.5; 1.12; and 1.95, and KS statistics are, from 0.89; 0.88; 0.76 and 0.69).

The results of the Kolmogorov-Smirnov test make sense because the distribution is random - whatever factor caused a value to be too high or too low affects only that value.

• Autocorrelation and Data Reliability Analysis

Autocorrelation data between GDP and independent variables helps in identifying lagged correlation and assessing data reliability (Table 6).

Table 6. Tests autocorrelations

Series: GDP
Lag	Autocorrelation	Std. Errora	Box-Ljung Statistic
			Value	df	Sig.b
1	-.253	.130	5.000	1	.025
2	-.127	.138	6.581	2	.037
3	.155	.140	8.843	3	.031
4	-.102	.143	9.944	4	.041
5	-.138	.144	11.911	5	.036
6	.077	.147	12.477	6	.052
7	-.052	.147	12.753	7	.078
8	-.007	.148	12.758	8	.120
9	.017	.148	12.787	9	.172
10	.041	.148	12.952	10	.226
11	-.128	.148	14.500	11	.207
12	-.046	.150	14.749	12	.255
13	.062	.150	15.274	13	.291
14	-.109	.150	16.857	14	.264
15	.176	.152	21.518	15	.121
16	.043	.155	21.753	16	.151

a. The underlying process assumed is MA with the order equal to the lag number minus one.

The Bartlett approximation is used.

b. Based on the asymptotic chi-square approximation.

The measurement of autocorrelation at Lag 1 and Lag 2 indicates that the correlations are statistically significant (p < 0.05), but negative and weak. From Lag 3 to Lag 16, the correlations are small and statistically insignificant (p > 0.05), suggesting that over longer periods, the correlation between GDP and the independent variables becomes negligible. Meanwhile, the Box-Ljung statistic shows that autocorrelation is prominent in shorter periods (p < 0.05 for Lag 1-5), but decreases for longer periods (p > 0.05 after Lag 5). Negative autocorrelations in the shorter periods may indicate external influences, while after Lag 3, the impact becomes insignificant, suggesting that the data is reliable for longer periods.

• Principal Component Analysis (PCA)

Table 7. KMO and Bartlett’s

KMO and Bartlett's Test
Kaiser-Meyer-Olkin Measure of Sampling Adequacy.		.287
Bartlett's Test of Sphericity	Approx. Chi-Square	303.387
	df	6
	Sig.	.000

The results of the Kaiser-Meyer-Olkin (KMO) test (Table 7) and Bartlett's Test of Sphericity indicate that PCA might be suitable for this data set. While the KMO value is 0.287 (below 0.5, suggesting poor suitability), the Bartlett's p-value (0.000) is very low, indicating that the variables are correlated and PCA can be used.

Table 8. Communalities

Variables	Initial	Extraction
Gross Domestic Product (GDP)	1.000	.734
Final Consumption (FCE)	1.000	.693
Exports (EXP)	1.000	.959
Imports (IMP)	1.000	.833
Extraction Method: Principal Component Analysis.

The Communalities results in Table 8 show that variables such as GDP, FCE, EXP, and IMP have high percentages of variance explained by the principal components, with extraction values ranging from 0.693 (FCE) to 0.959 (EXP), suggesting that all the variables contribute significantly to the formation of the principal components.

Table 9. Total variance explained

Component	Initial Eigenvalues			Extraction Sums of Squared Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	2.18	54.70	54.703	2.188	54.703	54.7
2	1.03	25.75	80.458	1.030	25.755	80.4
3	.756	18.90	99.361
4	.026	.639	100.00

Extraction Method: Principal Component Analysis.

The Total Variance (Table 9), Explained results show that the two main components explain over 80% of the total variance of the data, with the first component explaining 54.7% and the second 25.8%. The other components have a very small impact on the total variance.

10.png

Figure 10. Principal Component Analysis (PCA)

The results Scree plot or Figure 10, explained show that the first two principal components are important for explaining the variance in the data. PCA1 (2.18) explains the largest portion of variance and represents the most significant variable or combination of variables. PCA2 (1.03) still holds important information about variations among variables. PCA3 (0.756) and PCA4 (0.026) have very low values, especially PCA4, indicating minimal impact. Most of the dataset’s information can be reduced to PCA1 and PCA2, as the third and fourth components do not have a significant effect.

This result indicates that the PCA has successfully reduced the dimensionality of the data while preserving essential information.

The natural logarithm (Figure 11), transformation for the effect of imports on GDP, exports, and consumption helps improve economic analysis by reducing heteroskedasticity and clarifying the relationships between these variables. The graph shows imports on the horizontal axis, with values ranging from 35 to 71 Billion euro, and their effects on GDP, exports, and consumption on the vertical axis. This transformation aids in measuring elasticities and provides a clearer understanding of the economic impacts.

11.png

Figure 11. The impact of natural logarithm transformation on imports and economic variables

The Total Variance Explained results show that the two main components explain over 80% of the total variance of the data, with the first component explaining 54.7% and the second 25.8%. The other components have a very small impact on the total variance.

• Breusch - Pagan Test

Table 10. Breusch - Pagan test

In Table 10, a small chi-square value (along with an associated small p-value) indicates that the null hypothesis is true (i.e., the variances are all equal). In our case Chi-square χ²= 73.9 and p-value for Breusch-Pagan test: 0.1816. If the p-values of the coefficients are below a certain threshold (usually 0.05), this suggests that heteroscedasticity is present in the original model. Prob > chi², is the p-value corresponding to the Chi-Square test statistic, in this case, it is 0.181. Since this value is greater than 0.05, we cannot reject the null hypothesis and conclude that heteroscedasticity is not present in our data.

Table 13. Coefficients^a

• Second Modeli Empirik 2SLS

The testing of the first hypothesis using the Two-Stage Least Squares (2SLS) model.

H1.: The impact of imports, exports, and final consumption on annual GDP growth in Kosovo is positive and statistically significant.

H1.1.1: The impact of imports on annual GDP growth in Kosovo is positive and statistically significant.

The Two-Stage Least Squares (2SLS) model analyzes the impact of Imports (Log_IM), Exports (XM), and Final Consumption (Log_FC) on the growth of Log_GDP in Kosovo with a time lag (regression Eqs. (5)-(7)) is as follows (Table 14):

Table 14. 2SLS estimate of GDP and IM

• Impact Imports (Log_IM) on log GDP

2SLS Model: LogGDP_annual growth $=\widehat{a}+$ $\widehat{\beta} \widehat{\mathrm{X}}+\epsilon=\log \operatorname{GDP}=1.282+0.169 \mathrm{IM}$             (5)

Imports have a positive and statistically significant impact on GDP, with an effect of 16.9% for every 1% increase in imports. With R² = 0.576 and a p-value = 0.002, the model shows that imports explain a large portion of the variation in GDP. This highlights Kosovo's economy's dependency on imports for production and consumption. Research and Information System for Developing Countries, Prebisch [112] argued that developing countries rely heavily on imports and have difficulty developing their export industries [113].

• The Impact of Exports (Log_XM) on log GDP

H1.1.2: The impact of exports on annual GDP growth in Kosovo is positive and statistically significant.

2SLS Model:

$\begin{aligned} &{LogGDP} P_{{annual}} {growth}=\widehat{\mathrm{a}}+\widehat{\beta} \widehat{\mathrm{X}}+\epsilon=\text { LogGDP }=1.246+0.019 \mathrm{XM}\end{aligned}$           (6)

Exports have a positive but limited impact on GDP, where every 1% increase in exports leads to only a 1.9% increase in GDP (Table 15). With R² = 0.369 and a p-value = 0.005, the model is valid and statistically significant. This indicates that Kosovo's economy remains more dependent on imports than exports.

Table 15. 2SLS estimate of GDP and XM

• The Impact of Final Consumption (FC) on GDP

H1.1.3: The impact of Final Consumption on annual GDP growth in Kosovo is positive and statistically significant.

2SLS Model:

$\begin{aligned}{LogGDP}_{{annual}} & {growth}=\widehat{\mathrm{a}}+\widehat{\beta} \widehat{\mathrm{X}}+\epsilon =\operatorname{LogGDP} =1.890+0.703 \mathrm{FC}\end{aligned}$          (7)

Final consumption has the greatest impact on GDP, with an effect of 70.3% for every 1% increase (Table 16). With an R² of 57.6% and a p-value of 0.002, the model confirms its significance. This result indicates that Kosovo's economy relies mainly on domestic demand rather than exports.

Table 16. 2SLS Estimate of GDP and FC

According to these results, Kosovo's economy is characterized by a strong dependence on final consumption and imports, while exports play a secondary role. The 2SLS model analysis with lag for each variable (log) shows that final consumption (FC) has the largest impact on GDP, confirming that economic growth is primarily driven by domestic demand. Imports (IM) also significantly influence GDP, reflecting a high reliance on foreign goods and services. Meanwhile, exports (XM) have a smaller impact on GDP, which reflects the lack of a strong export sector in the economy. This economic reality makes Kosovo vulnerable to external factors and limits sustainable growth due to a weak export sector and high dependence on imports.

To improve the situation, Kosovo needs to increase production and exports, diversify its economy to reduce dependence on imports, and encourage investments. This would help shift from a consumption-based economy to a productive and export-oriented economy, ensuring more sustainable growth and reducing vulnerability to external shocks.

Verification of the first hypothesis (Table 17):

Table 17. Verification of the H1 - Kruskal-Wallis Test

The analysis shows that GDP varies significantly across FCE, EXP and IMP categories, as all tests result in a significance level of less than 0.05. This suggests that these factors have a significant impact on GDP, reinforcing the need for further analysis on the nature and extent of this impact.

• Log Transformation Effect on Dependent Variable

12.png

Figure 12. Legend information impact final consumption in GDP with time delay (2007-2023)

The histogram analyzes the relationship between imports, exports, consumption, and GDP. With a mean of 2.15 and a standard deviation of 4.465, the data shows significant variations from month to month. The vertical axis represents the frequencies, while the horizontal axis represents GDP. This graph helps in understanding the impact of economic factors on GDP, suggesting that there are substantial variations concerning the variables that influence Kosovo's GDP (Figure 12).

13.png

Figure 13. Legend information impact import in GDP with time delay (2007-2023)

The XGRAPH chart highlights the impact of imports on Kosovo's GDP from 2007 to 2023. The vertical axis, LagFC, represents final consumption expenditures, a key component driving GDP growth. Variations in LagFC (71.3 to 92.4) suggest a link between consumption and GDP growth both in the short and long term. Changes in LagGDP (from -1.4 to 5.77) show how past economic activity influences future GDP growth (Figure 13).

14.png

Figure 14. Legend information impact export in GDP with time delay (2007-2023)

Figure 14 provides insight into the impact of exports on Kosovo's GDP from 2007 to 2023, helping analyze aspects of economic growth acceleration and the effects of exports. Variations in LagXM (from 17 to 80.6) suggest a link between exports and GDP growth in both the short and long term. Changes in LagGDP (ranging from -11 to 5.77) indicate how past economic activities influence future GDP growth, suggesting that while exports may have an immediate impact on Kosovo’s economy, there may also be delays in their effect on GDP.

• Testing Hypothesis II, Pearson Correlation

H2: The relationship between annual GDP growth in Kosovo and the variables of imports, exports, and final consumption is statistically significant and positive.

Table 18 presents Spearman's rho correlation coefficients between four variables: GDP, Final Consumption Expenditure (FCE), Exports (EXP), and Imports (IMP). The significance level (Sig. 2-tailed) is also provided, indicating whether the correlation is statistically significant.

The results of the Spearman’s rho analysis for the Correlations between the variables Lag_GDP, LagFCE, LagIMP, and LagEXP reveal different relationships among these economic factors:

The correlation between Lag_GDP and LagFCE is 0.310, indicating a moderately weak but statistically significant positive relationship (p = 0.002), suggesting that an increase in GDP may stimulate growth in consumption expenditures. Meanwhile, Lag_GDP and LagIMP have a correlation of 0.583, showing a strong and statistically significant positive relationship (p = 0.000), confirming that GDP growth is closely linked to import growth, a common phenomenon in expanding economies. Finally, the correlation between Lag_GDP and LagEXP is 0.055, indicating a very weak and insignificant relationship (p = 0.589), suggesting that GDP has no notable impact on exports.

Table 18. Correlations

The correlation between LagFCE and LagIMP is -0.093, indicating a very weak negative and statistically insignificant relationship (p = 0.358), suggesting that final consumption expenditures have no significant impact on imports. Meanwhile, LagFCE and LagEXP have a correlation of -0.573, representing a moderately strong and statistically significant negative relationship (p = 0.000), where an increase in domestic consumption may reduce the demand for exported goods. Lastly, the correlation of 0.700 between LagIMP and LagEXP indicates a strong positive relationship, suggesting that an increase in imports may contribute to export growth through imported goods.

These results indicate that Kosovo’s economy has a strong relationship between GDP growth and imports, while the link between GDP and exports is very weak. The increase in consumption expenditures has a negative correlation with exports but a weak correlation with imports. Imports and exports are closely interconnected, reflecting the impact of international trade on Kosovo’s economy.

Kosovo's economy has shown rapid growth in recent years, primarily supported by the services sector, especially trade and tourism linked to the diaspora. However, industry and construction have contributed less to this growth. The increase in private consumption and fixed investments has helped support this development, while imports have had a negative impact, outpacing exports and widening the current account deficit. The study shows that imports and final consumption have a significant impact on GDP growth, while exports have a smaller effect.

The model most aligned with the findings of this study is a combination of Keynesian theory and the Prebisch-Singer hypothesis: i) Kosovo's economy is based on consumption and imports (Keynesian Model); ii) Exports do not play a major role in economic growth (Challenge to Ricardo, cite, Research and Information System for Developing Countries [112] argued that developing countries rely heavily on imports and have difficulty developing their export industries. iii) Kosovo is dependent on external sources such as remittances and imports (Dependency Theory). Kosovo relies on external sources like remittances and imports and needs policies that stimulate domestic production and export sector development to achieve sustainable economic growth.

This analysis strengthens the argument of this study for the need for balanced policies that foster investments and technological development to enhance Kosovo's economic structure. Lestari and Ignatia [79] also highlight the importance of policies that encourage exports through improved productivity and strategic investments. Additionally, Rudrappa and Veerabhadrappa [68] recommend that countries with small economies implement strategies to diversify their export markets in order to decrease reliance on imports and promote economic growth.

In this study, imports and final consumption have been identified as important factors for GDP growth, while exports have a smaller impact. This is in line with the Cobb-Douglas theory, which suggests that consumption (as a component of aggregate demand) and imports (as external factors influencing the domestic economy) can have a significant impact on economic growth. However, one feature of the Cobb-Douglas model is that production factors, such as capital and labor, have a more direct impact on production, and in this context, the study should have also considered other variables, such as the importance of investments in production and the development of domestic sectors, in order to achieve more sustainable and balanced growth.

In conclusion, addressing these structural challenges requires policy reforms that encourage innovation, domestic investments, and export-led growth. By reducing reliance on imports and leveraging its resources for competitive advantages, Kosovo can pave the way for more sustainable economic development.

6.1 Recommendations for Kosovo

i) Strengthening the Export Sector: Diversifying export products: Kosovo should focus on developing and diversifying its export products, especially in sectors such as agriculture, renewable energy, and IT, to increase export opportunities. Improving export infrastructure: Expanding trade opportunities and enhancing the quality of products can help boost competitiveness in international markets. Supporting small and medium-sized enterprises (SMEs): This sector has opportunities to increase exports of local products and create partnerships with international companies.

ii) Enhancing Domestic Competitiveness: Support for innovation and technology: Kosovo needs to invest in the development of new technologies to increase productivity and competitiveness in local industries. Improving investment conditions: Creating more favorable conditions for both foreign and local investors, including easing bureaucracy and providing fiscal and tax incentives. Developing workforce skills: Schools and higher education institutions should focus on developing new skills aligned with market demands, especially in fields that contribute to economic development, such as engineering and IT.

iii) Reducing Import Dependence: Encouraging domestic production: Kosovo could focus on strengthening domestic industries, especially in sectors with growth and competitive potential, such as agriculture and renewable energy production. Improving regional market linkages: Leveraging regional trade agreements to increase internal trade and boost export opportunities for Kosovar businesses.

iv) Investing in Human Resources and Technological Capacities: Support for research and development (R&D): Kosovo should create programs to support scientific research and technological development, providing opportunities for innovations that can contribute to the growth of various sectors. Improving access to finance for startups and innovators: By creating funds and opportunities to finance new ideas and product developments, Kosovo can benefit from modern technologies and create job opportunities.

v) Policies for Economic Stability and Reducing Consumption Dependence: Developing a sustainable growth model: Kosovo needs to develop policies that encourage sustainable investments, as well as include strategies that improve financial systems and create opportunities for revenue diversification. Increasing government transparency and accountability: Strengthening the role of institutions overseeing economic policy and increasing trust in institutions to encourage sustainable investments.

These recommendations are essential for developing a sustainable and balanced economy for Kosovo, increasing domestic competitiveness, and reducing dependence on external factors.

6.2 Study limitations

The potential limitations of this study include several important aspects that may affect the interpretation and application of the results. The use of a short time period (2007-2023) may not fully capture the impacts of global economic crises or changes in domestic policies that occurred before 2007 or after 2023. A longer period might have provided a more accurate picture of current developments. Additionally, while variables such as imports, final consumption, and exports are included, there may be other factors influencing GDP, such as foreign direct investments, technology, human resource development, or government policies, which were not considered in this study. Furthermore, although MLR and 2SLS models are useful for assessing the impact of various variables, they may not fully reflect the dynamic and complex relationships between these variables and GDP. A model like Fixed and Random Effects could have been a better choice for capturing these influences in more detail. Lastly, the study focuses primarily on the sectors of imports, exports, and final consumption, leaving out other significant sectors of the economy, such as services, industry, or technology, which may also have a substantial impact on GDP growth. These limitations may influence the accuracy and application of the study’s findings, but they also present opportunities for further research and the development of other models that could include a broader range of factors and longer time periods.