Decomposing drivers of transportation energy consumption and carbon dioxide emissions for the Philippines: the case of developing countries

Abstract

Global CO₂ emissions increased by 57.9% from 1990 to 2014, of which 21% is known to be from the transportation sector. In line with policy development, driving forces to energy consumption and emissions may be determined using decomposition analysis techniques. However, the detail of information required to perform such studies for the transportation sector in developing countries can be challenging. An attempt was made in this study to formulate a decomposition analysis framework considering data availability and limitation in developing countries. Furthermore, a suggestion of adjusting transport activity data using average oil price was proposed. An illustrative case study in the Philippines revealed that the most significant driver was transport activity, followed by energy intensity, and then population growth, which was both similar and contrary to all previous studies performed in developed and rapidly urbanizing countries, which pointed out to transport activity as the primary contributing force. For the Philippines, transport activity was an inhibiting force, whereas energy intensity was the primary contributing factor. The difference could be explained by the differences in mode shares and quality of life between countries. Looking at private vehicle ownership data, it is observed that growth rates are higher in the rural, than in the urban centers. Deriving from the findings, developing a comprehensive public transport plan is recommend for future growth areas, expansion and modernization of public transport services in the city, and strategic deployment of transport policies.

Keywords

transportation / LMDI / decomposition / developing country / emissions

Introduction

Global CO₂ emissions were 32.4 Gt in 2014, increasing by 57.9% since 1990 [¹]. This is alarming, considering that CO₂ is a major contributor to global warming and climate change. Of the global carbon emissions, 21% is originating from the transportation sector. The transportation sector is one of the largest consumers of fossil products. In fact, 6.57 billion tonnes of CO₂ were emitted by the sector in 2010 due to oil consumption alone. The International Energy Agency expects this to rise to 9.4 billion tonnes by 2035.

Post-Paris Agreement, plenty of developing economies still find it challenging to meet their targets. It is important to mention that the majority of emissions reductions in the recent decade have been from Kyoto parties with targets, which have already met or are scheduled to meet their committed reduction targets [¹]. 47% of committed parties had already met their targets before the end of 2014.

In line with policy development, driving forces to energy consumption and emissions may be determined using decomposition analysis techniques. Decomposition analysis techniques contextualize various contributing factors with regards to their relative effects on the energy consumption or emissions of a specific sector or economy. The essential factors considered in decomposition of energy consumption and emissions are population growth, economic activity, energy intensity, and carbon intensity, which are largely patterned from the Kaya identity [²].

The earliest decomposition analysis study on energy and emissions of the transportation sector was conducted in Ref. [³], studying freight and passenger transport in the United States. The factors considered were population growth, GDP, mode structure, and energy intensity. The increase in the population’s propensity to travel, reflected as GDP increase, was found as the most significant factor. In Ref. [⁴] a similar study was performed for Great Britain, considering population and total driving distance. Similar to the US study, activity was the most significant factor, while population showed insignificant contributions. On a larger scale, in Ref. [⁵] driving factors to transport CO₂ emissions of the EU14 was analyzed, considering the volume of travel and energy intensity. Similarly, the volume of travel, or activity, was shown to be the most significant contributor. From Asia, in Ref. [⁶] the energy intensity, mode share, number of trips, and travel distance for Japan were considered. The number of trips appeared to be the most significant factor contributing to transport energy consumption in Japan. Interestingly, a majority of recent work on the subject were from China. A total of papers decomposed driving forces of energy consumption and emissions for the transportation sector in China. In Ref. [⁷] a multi-regional analysis was performed for China, considering population, economic activity, energy intensity, energy structure, and emission factors. Significant disparity between regions was observed, and the main drivers were economic activity and population growth. Similar conclusions for China were obtained by Refs. [⁸‒¹³]. From an economy-wide perspective, in Refs. [¹⁴‒¹⁵] drivers to CO₂ emissions in the Philippines and Iran were unveiled, respectively, with emphasis on contributions of the transportation sector.

In summary, the most common contributing factors used in decomposition analysis studies for energy consumption and emissions of the transportation sector are listed in Table 1. Unlike other sectors, it is common for transportation studies to replace the population term with vehicle ownership data or number of trips. Transport mode share or vehicle technology mix is also an important consideration. For transport activity, it is common to be represented by travel distance or total kilometers traveled.

However, for developing countries, data availability is a problem. Travel distance data are not easily available, as much as annual transport mode share, and annual number of trips. This makes it difficult for scholars in developing countries to perform similar studies in the transportation sector. This is a problem, as plenty of growth is expected to occur in developing countries in the following decades, most especially in the Association of Southeast Asian Nations (ASEAN). The Southeast Asian economy is forecasted to grow by 5.1% in 2018, and similarly in the upcoming years [¹⁶]. Scholars from these contexts have to be equipped with the appropriate tools to perform similar studies in their own countries.

This paper provides a useful case study for scholars in developing contexts, by formulating the identity function for decomposition analysis considering data availability and limitation. The decomposition performed in this paper is anchored on the use of transportation sector value added (TVA) as a proxy for transport activity. Furthermore, a novelty of this paper is the use of average oil price to adjust transport activity data, which seems reasonable when an economic parameter (e.g. GDP) is used to represent transport activity.

Methods

Decomposition analysis

The preferred decomposition analysis method in this study is the logarithmic mean Divisia index (LMDI), which was introduced [¹⁷]. The technique involves the decomposition of a differential quantity, with its main advantages over other similar methods being not having residuals and having the capability to handle cases with zero values in the data set. The latter is important for energy-based studies, as it is normal to have zero demand for a certain type of fuel (e.g., renewable energy) in one year, and then have it in another year (e.g., introduction of renewable energy in a latter year). Readers are encouraged to see Refs. [¹⁸‒²⁰] for more information on the methodology.

The major expression in decomposition analysis is the identity function. The identity function is a prelude to the analysis, as it organizes all the contributing factors, or driving forces, to be considered in the analysis. The identity function utilized in this study considers data availability and limitation in developing countries, and is expressed in Eq. (1), while the deconstructed and abbreviated forms are expressed respectively in Eqs. (2) and (3).

{\mathrm{CO}}_2\mathrm{=population\times activity\times energy}\hspace{0.17em}\mathrm{intensity}

(1)\mathrm{\times fuel}\hspace{0.17em}\mathrm{structure\times emission}\hspace{0.17em}\mathrm{intensity,}

(2){\mathrm{CO}}_{\mathrm{2}}={\mathrm{\Sigma }}_i\mathrm{population}\times {\displaystyle \frac{\mathrm{transport}\mathrm{V}\mathrm{A}}{\mathrm{population}}}\times {\displaystyle \frac{\mathrm{energy}\hspace{0.17em}\mathrm{cons}.}{\mathrm{transport}\mathrm{VA}}}\times {\displaystyle \frac{{\mathrm{fuel}}_i}{\mathrm{energy}\hspace{0.17em}\mathrm{cons}.}}\times {\displaystyle \frac{{\mathrm{CO}}_{\mathrm{2}i}}{{\mathrm{fuel}}_i}},

(3){\mathrm{CO}}_2={\mathrm{\Sigma }}_i{\mathrm{pop\times act\times int\times str}}_i{\mathrm{\times emi}}_i.

Various forms of the identity function have already been used to study driving factors to energy consumption and carbon emissions of the transportation sector in different countries. The most common terms appearing in the identity functions are summarized in Table 1. For energy-related studies, the identity function usually follows the form I = PAT, or impact is equal to the product of population, affluence and technology. In this context, impact is energy consumption or any kind of emission. Population can be anything representing demand in general. It can simply represent human population, or sometimes, the number of vehicles also used in transport-related studies. Refs. [⁴] and [⁹] used human population data for this purpose, while Ref. [²¹] augmented their data using vehicle ownership per capita as well. Affluence represents the level of activity, or propensity to travel. On the other hand, technology is usually represented by multiple terms, such as energy intensity and carbon intensity. It is also common to see multiple structural terms, such as energy structure (i.e., energy mix) and economic structure (i.e., mix of sectors). For transport-related studies, it is also common to see a vehicle mix structural term (i.e., mode share).

With regards to calculating the individual effects or contributions of each driving factor, the LMDI technique is used, and the expression for each effect is shown from Eqs. (4) to (7). As can be seen from these equations, the effect of the emission intensity of each fuel type is not estimated as each is assumed to remain constant for the duration of the study period.

(4)\mathrm{\Delta }{C}_{\mathrm{pop}}^{\mathrm{T}}={\mathrm{\Sigma }}_i{\displaystyle \frac{c_i^{\mathrm{T}}-c_i^0}{\ln c_i^{\mathrm{T}}-\ln c_i^0}}\ln \left({\displaystyle \frac{P^{\mathrm{T}}}{P^{\mathrm{0}}}}\right),

(5)\mathrm{\Delta }{C}_{\mathrm{act}}^{\mathrm{T}}={\mathrm{\Sigma }}_i{\displaystyle \frac{c_i^{\mathrm{T}}-c_i^0}{\ln c_i^{\mathrm{T}}-\ln c_i^0}}\ln \left({\displaystyle \frac{G^{\mathrm{T}}}{G^{\mathrm{0}}}}\right),

(6)\mathrm{\Delta }{C}_{\mathrm{int}}^{\mathrm{T}}={\mathrm{\Sigma }}_i{\displaystyle \frac{c_i^{\mathrm{T}}-c_i^0}{\ln c_i^{\mathrm{T}}-\ln c_i^0}}\ln \left({\displaystyle \frac{I^{\mathrm{T}}}{I^{\mathrm{0}}}}\right),

(7)\mathrm{\Delta }{C}_{\mathrm{str}}^{\mathrm{T}}={\mathrm{\Sigma }}_i{\displaystyle \frac{c_i^{\mathrm{T}}-c_i^0}{\ln c_i^{\mathrm{T}}-\ln c_i^0}}\ln \left({\displaystyle \frac{S_i^{\mathrm{T}}}{S_i^0}}\right),

where C is CO₂ emissions (tCO₂); T and 0 refer to final and base year, respectively; i refers to fuel type i; pop refers to population effect; act refers to activity effect; int refers to energy intensity effect; str refers to fuel structure effect; while P, G, I, and S refer to population (number of persons), transport activity (transport value added in USD per capita), energy intensity (ktoe per USD transport value added), and fuel mix (%), respectively.

Driving factors

To represent the activity in this study, the transport sector value added (TVA) is used. Though total vehicle-kilometers is the ideal measure for transport activity, this kind of data are not often easily available in developing countries, unlike TVA which is available in most countries publishing statistical yearbooks. The TVA is the economic contribution of the transportation sector to the national gross domestic product (GDP). Citing a few studies which used transport GDP as a proxy for transport activity before, Ref. [⁷] used it to uncover driving forces to transportation sector CO₂ emissions in multiple Chinese regions, while Ref. [³] used it to perform a similar analysis for the United States. In comparison to economy-wide GDP, using TVA provides some selectivity, which makes it possible to filter the portion which is 100% related to transport activity. There are plenty of sectors in the economy whose revenue generating mechanisms are not heavily dependent on transportation (e.g., education, healthcare, and manufacturing).

According to Ref. [²²], TVA is the economic output of all for-hire transportation services in the country, minus the intermediate inputs to it (i.e., fuel and equipment costs). The fact that it does not cover in-house transportation expenses of various industries and private transport demand from households is a weakness. However, these are covered by the energy consumption term in this study.

After population and transport activity, energy intensity is the ratio between transport energy consumption and TVA. The energy consumption data available for the study covers all public and private modes, including land, sea, rail and air, and for all purposes (i.e., household, private industry, freight, etc). Therefore, even though transport activity only covers for-hire transport services, the emissions and energy consumption estimated are for the whole transportation sector. Should private vehicle transport demand rise, this would be reflected as an increase in energy intensity of the transportation sector. The reason for this is that the energy intensity term would increase without a proportionate growth from the activity term. To confirm this, mode share from the two year data available (i.e., 1995 and 2014) will be compared, together with vehicle registration data from recent years in the next section.

In addition, the fuel structure (i.e., fuel mix) of the transportation sector is also included in the analysis. This is called the structural effect. It simply reflects the changes in CO₂ emissions caused by a sudden shift to less or more carbon intensive energy sources in the transportation sector. Moreover, the carbon intensity of each fuel type is shown to be included in the identity function (see Eq. (2)), however, the values are assumed to be constant for the duration of the study period.

Adjustment of transport value added

A novelty of this study is the use of average oil price to adjust the value of TVA. To consider the effects of fluctuating oil prices to the value of transportation sector output, TVA is divided with average fuel price, as expressed for each year, k, in Eq. (8). This is done for each year in the data set. Both TVA and average fuel price are taken in constant 2000 USD to account for the effects of inflation. The annual average fuel price and TVA from 2000 to 2014 are shown in Fig. 1.

(8)\mathrm{Adjusted}\hspace{0.17em}{\mathrm{TVA}}_k={\displaystyle \frac{{\mathrm{TVA}}_k}{\mathrm{Ave}\mathrm{.oil}\hspace{0.17em}{\mathrm{price}}_k}}.

Data sources

Population and TVA data used in this study were taken from the Philippine Statistical Yearbook [²³]. The energy mix for electricity generation was derived from the Philippine Department of Energy [²⁴], while the fuel mix of the Philippine transportation sector was obtained from the International Energy Agency [²⁵]. The carbon intensities (i.e., emission factors) of the various fuels were taken from various references. For mobile combustion, emission factors were derived from Vol. 2, Ch. 3 of the 2006 Intergovernmental Panel on Climate Change Guidelines for National Greenhouse Gas Inventories [²⁶]. For stationary combustion (i.e., electricity generation), emission factors were derived from Vol. 2, Ch. 2 of the 2006 Intergovernmental Panel on Climate Change Guidelines for National Greenhouse Gas Inventories [²⁷], the Environmental Protection Agency [²⁸,²⁹]. For adjusting TVA, annual average oil prices were obtained from Ref. [³⁰].

Results and discussion

Comparison of drivers using adjusted and unadjusted TVA

The results will be discussed by comparing the drivers resulting from the use of unadjusted and adjusted TVA. Because the drivers are calculated from the actual historical energy use and emissions data, it is impossible to get different net effects between the two, and only the balance of effects can change. Primarily, it is observed that the effects are magnified in the adjusted TVA scenario compared to the unadjusted one. For example, using adjusted TVA, the largest effects reach up to 20 million tons CO₂, while it is only approximately 5 million tons CO₂ (see Fig. 2) using the other. Looking at the raw data for TVA, it can be understood that the reason for this is that the annual fluctuations found in the adjusted TVA are larger than those in the unadjusted data. The reason for this is more methodological – since decomposition is somewhat a form of sensitivity analysis, larger effects are assigned to activity to reflect these fluctuations.

Subsequently, there is a major difference with regards to the primary driving force observed, which is population effect when an unadjusted TVA is used, and is energy intensity when an adjusted TVA is used.

Looking at both TVA’s, the unadjusted values appear constant over an extended period, and only exhibit an increasing trend during the final four years of the period. On the other hand, after adjusting TVA using the average fuel price, a trend is revealed where activity initially grows, then goes on a decreasing trend until 2006, and somewhat stays constant for the next few years thereafter (see Fig. 3). Because of this, year-on-year fluctuations by transport activity become larger and more significant than that of population and energy intensity, and thus become a stronger explanatory factor for the changes in CO₂ emissions from the sector.

Population effect remains to be a significant contributing factor in the growth of CO₂ emissions even when an adjusted TVA is used, however, its contribution is overshadowed by economic activity and energy intensity. In relation to that, it is interesting to observe the reversal of the role of energy intensity. Using an unadjusted TVA, energy intensity is seen as an inhibiting factor, whereas it becomes a contributing factor after adjusting the TVA data. Since energy intensity is the ratio between energy consumption and TVA, the raw data for energy intensity is directly affected by the adjustment of TVA. A key factor is the significant decrease in transport activity in the middle of the period. Though activity remained steady in the latter half, it never really manifested a strong increasing trend. Though population was increasing, it was not high enough to explain the high CO₂ emissions. To keep the emissions at the same level even with a significant drop in activity, the energy intensity or the efficiency of energy use must be increasing – and that is what is observed on the decomposition analysis using adjusted TVA. Taking note of the differences between the two, the ensuing discussions will be using the results from the adjusted TVA calculations.

Analysis of drivers (annual and aggregated)

CO₂ emissions from the transportation sector in the Philippines, interestingly, did not grow significantly from 2000 to 2014. The sector steadily emitted approximately 26 MtCO₂ per year in the period. However, it is observed that the study period ended signifying an increasing trend of CO₂ emissions, which could be concerning.

From an aggregated perspective, transport activity is observed to be the most significant (inhibiting) factor, followed by energy intensity (contributing), and then population growth (contributing). The effects of changes in energy structure and emission factors appear to be insignificant. With regards to the importance of the activity effect in transportation sector CO₂ emissions, the findings are in congruence with previous studies [¹⁰,¹³], but just having the opposite effect. The activity effect has been observed as a contributing factor in all previous studies, which happen to be limited to developed and rapidly urbanizing (i.e., China) countries. To the knowledge of the authors, the present study is the first study on the transportation sector from a developing country.

Looking at the annual effects (Fig. 4), transport activity and energy intensity almost cancel each other out. Also, the annual effects of population growth are too small compared to transport activity and energy intensity to be noticed. The fluctuations in oil price appear to significantly influence transport CO₂ emissions in the country. For example, though the value of unadjusted TVA (see Fig. 1) only increased slightly from 2000 to 2001, fuel price decreased significantly in the same period – meaning in fuel terms, the amount of fuel consumed actually increased by a lot. This is reflected by a large positive contribution by activity effect from 2000 to 2001. In the years thereafter, fuel price increased almost exponentially while TVA (unadjusted) was steady at around 2.4 billion USD per year – again, in fuel terms, this means that fuel consumption decreased significantly in the period, as signified by the inhibiting effect of transport activity from 2001 to 2008. Moreover, fuel price rose to its highest value in the years as the study period ended, but TVA still managed to obtain a steady increasing trend. This resulted in transport activity becoming a contributing factor as the study period ended. Strong national economic growth toward the end of the period (see Fig. 4) caused increased spending in transport. The more vibrant economy resulted in increased transport activity, and possibly more private transport utilization. Using panel data from 24 Chinese provinces, Ref. [³¹] showed that transport facilities were key differentiating factors for economic growth. Ref. [³²] also mentioned that transport investments were seen as engines for economic growth and development in the United States.

With that note, energy intensity is generally portrayed as a contributing factor in decomposition analysis. As explained in the previous section, the fact that transport CO₂ emissions maintain their high levels amidst significant decreases in transport activity could only mean that energy intensity is increasing. Energy intensity is calculated as energy consumption per TVA, meaning it can also be interpreted as energy efficiency. A potential explanation for the increasing trend of energy intensity is the shift to more energy intensive transport modes, which is analyzed more thoroughly in Subsection 3.3.

Critical analysis of results

In connection to the general analyses above, two main themes are present: decreasing transport activity, and increasing energy intensity. It is evident that transport activity is having an increasing trend toward the end of the study period.

Figure 5 puts together various plots showing possible correlations between different parameters involved in the analysis of drivers. The top portion of Fig. 4 shows national GDP and TVA, both of which are expressed in constant year 2000 USD. The trend indicates that both of them are on an increasing trend. On the other hand, the bottom portion of Fig. 4 compares TVA and adjusted TVA using fuel price. The plot exhibits an interesting contrast between the two and presents the significant effect of normalization on the activity data. Since fuel price has been constantly increasing at a higher rate than TVA, the actual value of transport activity seems to be rather decreasing every year. It is important to take note though that it is showing some recovery toward the end of the study period.

Moreover, the charts in Fig. 5 show the relationship between energy intensity, population, and transport activity. Population growth and energy intensity manifest the same increasing trend. The contrasting relationship between energy intensity and transport activity is interesting, as energy intensity continues to increase while transport activity decreases. Primarily, it is surprising to observe such a trend for transport activity, but it presents an interesting opportunity for further investigation. As reiterated in Subsection 3.2, since transport energy consumption does not decrease as much as transport activity does in the study period, it means that the efficiency of energy use in the transportation sector is declining. For both trends to happen simultaneously, a possible reason would be an increase in more energy intensive modes (e.g. private car use), combined with a larger part of the population not being able to afford high transportation costs. This is a potential issue of energy equity – a case where a smaller portion of the population is responsible with majority of energy use. This scenario is extreme, but not impossible in developing countries, where the gap between rich and poor is very wide.

Finally, the bottom-right chart in Fig. 5 puts together energy intensity, national GDP, and population growth, which altogether demonstrate an increasing trend. As the economy surges forward in a developing country, income levels rise, and the danger of increasing energy intensity through more private car use poses a threat to energy security and emissions reduction targets.

In relation to the observed increase in energy intensity of the transportation sector, it is possible to compare mode share in the capital city of Metro Manila from Refs. [³³,³⁴]. Metro Manila is a megacity of approximately 12 million people, and is contributing 38.13% of the national GDP [³⁵]. The mode share comparison is displayed in Fig. 6. Jeepneys are minibuses (mostly open-air) which serve as the primary mode of public transport in the Philippines. Each jeepney can seat 14 to 20 people, and follow fixed routes. Tricycles are motorcycles equipped with a side car servicing designated zones, and can cover short to medium trip distances. Pedicabs are about the same, but operate with bicycles instead of motorcycles. UV/HOV’s, or utility vans, function the same way as jeepneys.

Interestingly, car use did not increase between 1995 and 2014 in Metro Manila. However, it is evident that the use of motorcycles significantly grew in 2014. The UV/HOV and Pedicab data were not as widely utilized in 1995, and hence the data on them were not yet available in 1995. Also notable is the increase in walking trips in the city.

Though the motorcycle counts as private transport, it cannot explain the significant increase in transport energy intensity of the country by itself. It is worthwhile to look further outside Metro Manila and analyze car ownership data, but unfortunately, mode share information is not directly available outside Metro Manila. Instead, using the light duty vehicle (LDV) registration data from Ref. [³⁶], it is observed that the growth rates of car ownership in the provinces outside Metro Manila are significantly higher. The data suggests that a majority of the increase in transport energy intensity is happening outside the megacity. Though major cities are also developing outside Metro Manila (e.g., Angeles City in Region 3, Cebu City in Region 7, and Davao City in Region 11), most of the regions can be considered rural. Figure 7 compares LDV ownership growth rates between Metro Manila and outside regions from 2000 to 2015.

Conclusions and policy recommendations

In this paper, drivers to changes in transport CO₂ emissions from a developing country – the Philippines were analyzed. As observed in related literature, both transport activity and energy intensity play an important role. However, a role reversal was seen for transport activity in this paper. It was reported to be a contributing factor in all previous studies, but it was observed to be an inhibiting factor in this paper. Since all previous studies to date have been from developed and rapidly urbanizing countries (i.e., China), and the Philippines happens to be a developing country, it is worth investigating if the same trend can be found in other developing countries so that the underlying reason can be studied further. Furthermore, it would be interesting to test the adjustment method for TVA used in this paper on the data of developed countries, and observe if the trend for transport activity would also be reversed. The decreasing trend in transport activity resulted from the adjustment of transport value added data using the annual average fuel price. It is also possible that what is observed in the study in this paper is simply the beginning of increased transport activity in the Philippines – a hint of which is evident already from the latter end of the study period (i.e., 2011 onwards).

With regards to the use of fuel price-adjusted TVA as a proxy for transport activity, two major assumptions need to be considered. First, it is assumed that transport activity is directly proportional to fuel consumption. Significant discrepancies could arise if non-fuel consuming modes become significant in the mix. Second, it is assumed that transport activity is dominated by petroleum-based modes. However, this can be easily addressed if an average energy cost weighted by fuel mix is used instead of average fuel price (per barrel).

As for energy intensity, it is seen as a contributing factor to increases in transport CO₂ emissions. It is believed by the authors of this paper that this is due to increased private transport use, which is observed to occur outside the country’s financial center of Metro Manila. Vehicle registration data show that light duty vehicle ownership has been growing faster in rural provinces than in urban Metro Manila. This should serve as a precaution, as Philippine GDP growth rates have been the highest in recent years [³⁷], and household incomes are expected even more to rise. Together with the recent implementation of a new law increasing excise taxes for new vehicle purchases [³⁸], it would be interesting to find out how these developments will interact with transport energy intensity and CO₂ emissions.

With this, the potential policy recommendations arising from the findings in this study were discussed. The policies are conceptualized particularly for the Philippines, but should be applicable to other developing countries as well.

(1) Development of a comprehensive public transport plan in future growth areas outside the urban centers

Based on new mode share and LDV growth rate data, much of the future growth in transport activity appear to happen outside the present urban centers. With this, emphasis on provincial/rural public transport development is necessary to mitigate private vehicle ownership growth. Initially, there should be efforts to identify future growth areas and development corridors. Visionary policy making can prevent future urban areas from suffering the similar fate of current cities suffering from severe private transport reliance, pollution, and constrained public transport expansion.

(2) Expansion and modernization of public mass transport services

A passive solution to reducing private transport use is to increase the utility value of public transport modes. As income levels rise, cheaper fares would no longer be enough to attract passengers. Increased reliability, efficiency, convenience, and comfort can significantly increase public transport use and take away people from their cars. Expansion of services to peripheral and hard to reach areas, as well as modernization of safety and maintenance features of public transport equipment are also necessary. Developing countries need to invest in revitalizing their existing public transport systems from its poor state, to encourage the population to use them.

(3) Investment in standards

As part of modernization, the governments of developing countries also need to start investing in developing and improving their transportation standards. For example, the Philippines have recently announced a nationwide modernization of public utility jeepneys, which primarily focuses on addressing the violations of old jeepney design on the UNECE standards. Also, there should be a policy to phase out old vehicle models, due to their poor fuel efficiency and combustion, producing harmful pollutants to the atmosphere. Moreover, a simple adoption of up-to-date fuel standards (e.g. Euro standards) is important to stay at par with global clean fuel initiatives.

(4) Strategic and timely deployment of policies

In as much as prohibitive policies are necessary to control private vehicle use, it is believed by the authors of this paper that the development of public transport systems in developing countries is really the single, most important issue. As public transport systems are usually poor in developing countries, a strategic and timely deployment of policies is recommended. That is, efforts on making public transport desirable and available first should be strengthened, before penalizing current private transport users. Before discouraging or forcing people out of their current routine, there has to be efforts first to check if the alternative is already acceptable and sufficient to cater to the additional demand. Sudden implementation of a more restrictive number coding system, for example, to limit private LDVs on the road would spike demand for public transit. If the public alternatives are not ready to supply the induced demand by this policy, the policy should not be deployed yet. The same effect is experienced when a stricter quality control procedure is applied to public utility jeepneys.

To conclude, the authors of this paper recommend that similar studies be conducted in other developing countries to provide further validation of our findings. The current research literature on transport energy use and emissions would also benefit a lot from insights in developing country contexts. Besides, there is a strong urgency on this issue, given the projected population and economic growth rates in the developing regions of the world, most especially in Southeast Asia.

References

[1] International Energy Agency. CO₂ Emissions from Fuel Combustion: Highlights, 2016 ed. Paris, France, 2016

[2] Kaya Y. Impact of carbon dioxide emission control on GNP growth: interpretation of proposed scenarios. In: Proceedings of the IPCC Energy and Industry Subgroup, Response Strategies Working Group. Paris, France, 1990

[3] Lakshmanan T R, Han X. Factors underlying transportation CO₂ emissions in the USA: a decomposition analysis. Transportation Research Part D, Transport and Environment, 1997, 2(1): 1–15

[4] Kwon T. Decomposition of factors determining the trend of CO₂ emissions from car travel in Great Britain (1970‒2000). Ecological Economics, 2005, 53(2): 261–275

[5] Papagiannaki K, Leontarakis G, Diakoulaki D. Decomposition analysis of CO₂ emissions from road transport in EU15. In: Proceedings of the 10th International Conference on Environmental Science and Technology. Kos Island, Greece, 2007

[6] Jian J. A factor decomposition analysis of transportation energy consumption and related policy implications. International Association of Traffic and Safety Sciences, 2015, 38: 142–148

[7] Guo B, Geng Y, Franke B, Hao H, Liu Y, Chiu A. Uncovering China’s transport CO₂ emission patterns at the regional level. Energy Policy, 2014, 74: 134–146

[8] Wang W W, Zhang M, Zhou M. Using LMDI method to analyze transport sector CO₂ emissions in China. Energy, 2011, 36(10): 5909–5915

[9] Wang Y, Hayashi Y, Kato H, Liu C. Decomposition analysis of CO₂ emissions increase from the passenger transport sector in Shanghai, China. International Journal of Urban Sciences, 2011, 15(2): 121–136

[10] Zhang M, Li H, Zhou M, Mu H. Decomposition analysis of energy consumption in Chinese transportation sector. Applied Energy, 2011, 88(6): 2279–2285

[11] Ding J, Jin F, Li Y, Wang J. Analysis of transportation carbon emissions and its potential for reduction in China. Chinese Journal of Population Resources and Environment, 2013, 11(1): 17–25

[12] Wu H, Xu W. Cargo transport energy consumption factors analysis: based on LMDI decomposition technique. In: Proceedings of 2014 International Conference on Environment Systems Science and Engineering. IERI Procedia, 2014, 9: 168–175

[13] Fan R, Lei Y. Decomposition analysis of energy-related carbon emissions from the transportation sector in Beijing. Transportation Research Part D, Transport and Environment, 2016, 42: 135–145

[14] Sumabat A K, Lopez N S, Yu K D, Hao H, Li R, Geng Y, Chiu A S F. Decomposition analysis of Philippine CO₂ emissions from fuel combustion and electricity generation. Applied Energy, 2016, 164: 795–804

[15] Mousavi B, Lopez N S A, Biona J B M, Chiu A S F, Blesl M. Driving forces of Iran’s CO₂ emissions from energy consumption: an LMDI decomposition approach. Applied Energy, 2017, 206: 804–814

[16] Asian Development Bank. Asian Development Outlook 2017 Update. 2017–12–28,

[17] Ang B W, Zhang F Q, Choi K. Factorizing changes in energy and environmental indicators through decomposition. Energy, 1998, 23(6): 489–495

[18] Ang B W. The LMDI approach to decomposition analysis: a practical guide. Energy Policy, 2005, 33(7): 867–871

[19] Ang B W, Liu N. Energy decomposition analysis: IEA model versus other methods. Energy Policy, 2007, 35(3): 1426–1432

[20] Ang B W. LMDI decomposition approach: a guide for implementation. Energy Policy, 2015, 86: 233–238

[21] Papagiannaki K, Diakoulaki D. Decomposition analysis of CO₂ emissions from passenger cars: the cases of Greece and Denmark. Energy Policy, 2009, 37(8): 3259–3267

[22] Bureau of Transportation Statistics. Transportation’s contribution to the economy. 2016,

[23] Philippine Statistics Authority. 2015 Philippine Statistical Yearbook. 2015,

[24] Philippine Department of Energy. Power statistics. 2015–10–08,

[25] International Energy Agency. 2017 IEA statistics report. 2017–12–28,

[26] Intergovernmental Panel on Climate Change (IPCC). 2006 guidelines for national greenhouse gas inventories: mobile combustion. 2006,

[27] Intergovernmental Panel on Climate Change (IPCC). 2006 guidelines for national greenhouse gas inventories: stationary combustion. 2006,

[28] Environmental Protection Agency (EPA). Emission factors for greenhouse gas inventories. 2014,

[29] DiPippo R. Geothermal Power Plants: Principles, Applications, Case Studies and Environmental Impact, 3rd ed. Oxford: Elsevier, 2012

[30] Association of the German Petroleum Industry (MWV). Average annual OPEC crude oil price from 1960 to 2018 (in U.S. Dollars per Barrel). 2018–02–27,

[31] Démurger S. Infrastructure development and economic growth: an explanation for regional disparities in China? Journal of Comparative Economics, 2001, 29(1): 95–117

[32] Deakin E. Sustainable development and sustainable transportation: strategies for economic prosperity, environmental quality, and equity. 2001,

[33] Metro Manila Urban Transportation Integration Study (MMUTIS). Metro Manila urban transportation integration study: technical report No. 4. —Transportation demand characteristics based on person trip survey. Japan International Cooperation Agency (JICA), 1999

[34] ALMEC Corporation. Oriental Consultants Global Co., Ltd. The project for capacity development on transportation planning and database management in the Republic of the Philippines, MMUTIS update and capacity enhancement project (MUCEP): transportation demand characteristics based on MUCEP person trip survey. Japan International Cooperation Agency (JICA), 2015

[35] Philippine Statistics Authority. Gross regional domestic product. 2018–03–14,

[36] Land Transportation Office. Registered motor vehicles by classification and region. 2018–03–20,

[37] Bloomberg. Philippine GDP growth of 6.9% beats all estimates. 2018–03–14,

[38] Department of Finance, Philippines. Tax reform for acceleration and inclusion. 2018–03–14,

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature