Household Water Consumption in Spain: Disparities between Region
Round 1
Reviewer 1 Report
I read the paper with interest and appreciated a lot the use of Phillips-Sul approach as a way to detect convergence in water consumption patterns across regions.
The introduction could be enlarged a little bit by citing papers that have delt with water consumption and the achievement of SDGs (see Hutton & Varughese, 2016; Hoekstra et al., 2017; Ho et al., 2020)
The methodological approach, its justification and presentation are fine. I also found reasonable the second-step that seeks to delve into the determinants of regions' belonging to clubs. Regarding this latter empirical exercise, I would like to suggest adding two variables to the regression model (one each on environmental conditions and household structure): i) water scarcity (see Garrone et al., 2019) ii) average household size to account for economies of scale in the use of water at household level (see Marzano et al., 2018).
Reference
Garrone, P., Grilli, L., & Marzano, R. (2019). Price elasticity of water demand considering scarcity and attitudes. Utilities Policy, 59, 100927.
Ho, L., Alonso, A., Forio, M. A. E., Vanclooster, M., & Goethals, P. L. (2020). Water research in support of the Sustainable Development Goal 6: A case study in Belgium. Journal of Cleaner Production, 277, 124082.
Hoekstra, A. Y., Chapagain, A. K., & Van Oel, P. R. (2017). Advancing water footprint assessment research: Challenges in monitoring progress towards sustainable development goal 6. Water, 9(6), 438.
Hutton, G., & Varughese, M. (2016). The costs of meeting the 2030 sustainable development goal targets on drinking water, sanitation, and hygiene. World Bank.
Marzano, R., Rouge, C., Garrone, P., Grilli, L., Harou, J. J., & Pulido-Velazquez, M. (2018). Determinants of the price response to residential water tariffs: Meta-analysis and beyond. Environmental Modelling & Software, 101, 236-248.
Author Response
Referee 1
We would like to express our gratitude to the referee for the useful comments and suggestions that have helped us to improve the quality of our work. We have tried to reflect them in the new version.
We provide below a detailed answer to your questions:
- Following your suggestions, we have enlarged the introduction, citing the paper you recommended. It can be seen in lines 55-56 and 69-79.
- We agree with your view that the variables you mention are of great interest. In fact, we included the household size in our set of potential explanatory variables. However, its discriminating value is not sufficient to appear in the final specification. Following your point, we have recognized the importance of these two variables as explanatory variables in lines 368-374 and 382-384. Additionally, we have included the average values of the household size in the descriptive analysis (see the last rows of Table 3).
- We have included all the references you have recommended.
Reviewer 2 Report
The manuscript "Household water consumption in Spain. Disparities between Regions" is an interesting paper in line with the aim of the journal. It is well written and organized, but it needs to be improved before publication.
In the introduction section the literature review needs to be deeper investigated by considering other studies carried out on the specific field. A table summarizing these studies could be interesting. Moreover, the scientific gap needs to be better highlighted as well as the novelty of the manuscript.
In Table 1: the caption needs a better description and the units of measurement need to be written in the table. The considered values in columns 2000 and 2018 are average values? The Min and Max is for the whole series (2000-2018)? The authors should better specify all of this details and information.
The total water consumption reduction nearly observed was due to new habits of the people, the use of sustainable devices to reduce consumption (i.e. flow reducer, ect) or? The authors should discuss the reduction observed.
Overall a flowchart that summarizes the whole procedure is suggested.
A comparison with the trends in terms of water consumption carried out in other studies in Europe is suggested, especially based on the recent pandemic situation.
The conclusion section is too long; it needs to be reduced and the main findings highlighted by a bullet list. Moreover, the limit of the research and future investigations needs to be written.
English language and style needs to be revised.
Author Response
Referee 2
We would like to express our gratitude to the referee for the useful comments and suggestions that have helped us to improve the quality of our work. We have tried to reflect them in the new version.
We provide below a detailed answer to your questions:
- Following your suggestions, we have added a new table (appendix 1) that includes relevant papers from the household water consumption literature. We have also included some of the most recent reviews of this previously published literature.
- We have also tried to highlight the importance of our results. This can be seen in lines 105-106 and 113-126.
- We have improved Table 1 including the description of the units of measurement and providing more information about the figures included in the different columns.
- This is a very important point that would deserve a full analysis. However, the techniques employed in this paper do not allow for determining what is the source of the reduction of the water consumption, although they can help us to observe whether the evolution has been similar across the Spanish regions and what are the factors that can better explain these differences. We have included a footnote on page 4 where we discuss this briefly, providing useful references that focus on this issue in depth.
- We have discarded the flowchart you mention. We are not familiar with the software necessary and, therefore, we do not believe that we can generate a flowchart good enough to be added to the paper.
- This is another very important point. In fact, we look forward to using the post pandemic data to analyze them. However, these data are not available for Spain yet. Even worse, we have not been able to obtain a homogenous European database on water consumption to carry out a similar analysis to the one reported in this paper but focused on Europe. We greatly appreciate your suggestion, but we cannot follow it due to the lack of data.
- We have improved the conclusion section. This can be seen in lines 567-571.
- We have edited the text with the help of a native English teacher.
Reviewer 3 Report
Household water consumption in Spain. Disparities between Regions
This paper addresses a very important question using an innovative technique in the form of a longitudinal analysis rather than the common cross-sectional form and produces some intriguing results. That analysis also yields a high explanatory power although R2 is not a measure of statistical fit. Sustainable urban water management requires that we drive down household demand to sustainable levels so we need to discover how to achieve this: what are the effective incentives to induce a reduction. At the same time, we know that in some parts of Europe, notably Copenhagen and Germany, there have been dramatic reductions in average household consumption in recent years but we do not know why this happened or probably more importantly which components of household demand have been reduced and by how much.
There are some editing changes which I feel would help the reader:
- How many water and wastewater utilities there are in Spain? Do all set a uniform price regime across their service area?
- The term ‘club’ has a very specific meaning in economics, a ‘club good’ being different to both a public good and a private good. Wouldn’t ‘cluster’ or ‘group’ work as well in the context of this paper?
- In most European countries, it is only buildings that are metered, not individual dwellings. I see that over 66% of the people in Spain live in apartments. Are these all individually metered? Equally, I would expect to see a difference in average water consumption between apartment dwellers and house dwellers. If there is a difference in the proportions of apartments in different regions, then this could be a potential independent variable. Otherwise the included independent variables are sensible.
- The paper has a nice appendix giving the data sources. I agree that the average price should include both the potable water service and the foul sewage service. Does the data source however include stormwater drainage, something which has no necessary linkage to the other two services?
- Again, on the appendix, does the source give any details of the method by which the figure of average water consumption was calculated or provided? The analysis hangs on these numbers being reliable and from a consistent methodology.
I would like to see some more discussion on the methodological problems that had to be addressed. The study is based on averages from aggregate data. This first raises the risk of the ecological fallacy, where differences between aggregates are interpreted to account for differences with each aggregate. A second problem is the problem of regression to the mean, we lose the variance within the population and equally lack knowledge of the original frequency distribution. A further issue in regard to the average price is that unless the price regime is one of a standard unit rate per unit volume, without any standing charge, then the average price per unit varies with quantity consumed. Quite often, the average price per unit shows a U-shaped distribution; a fixed standing charge on the one hand and a rising block charge on the other.
Another country where there are quite marked differences in household water consumption between regions is Germany. Some years ago, I tried a cross-sectional analysis using various potential explanatory variables. I did not find the resulting multiple regression results very convincing partly because of the above reasons.
As Valencia is the city I have visited most often in recent years, I was fascinated to see that it shows up in Group 1 with high per capita water consumption. A paper I found in consequence was:
- Maldonado-Devis et al 2021 “Panel Data Estimation of Domestic Water Demand with IRT Tariff Structure: The Case of the City o Valencia (Spain)”, Sustainability 13, 1414
This found significantly differences within the city region itself.
The paper includes a very comprehensive set of useful references
So whilst I would like to see some small editing changes, a very publishable paper.
Author Response
Referee 3
We would like to express our gratitude to the referee for the useful comments and suggestions that have helped us to improve the quality of our work. We have tried to reflect them in the new version.
We provide below a detailed answer to your questions:
- How many water and wastewater utilities there are in Spain? Do all set a uniform price regime across their service area?
We have revised the availability of this data at the Spanish Statistic Institute (INE), but we have not been able to find the answer to your questions. Having said that, we agree with the referee that these variables could help us to better understand the disparities in the water consumption reduction because, as we mention in the paper, increasing efficiency may have played an extremely important role in this regard. Unfortunately, we cannot test for it due to the mentioned lack of data.
- The term ‘club’ has a very specific meaning in economics, a ‘club good’ being different to both a public good and a private good. Wouldn’t ‘cluster’ or ‘group’ work as well in the context of this paper?
The methodology developed by Phillips and Sul (2008) uses the notion of convergence club for those convergent regions that exhibit similar behavior. Then, we prefer to maintain the original concept of convergence club. In any event, and following your suggestion, we have employed the term “convergence clubs” throughout the paper (instead of “the club”), although we have maintained the denomination of the three estimated convergence clubs: Club 1, Club 2 and Club 3.
- In most European countries, it is only buildings that are metered, not individual dwellings. I see that over 66% of the people in Spain live in apartments. Are these all individually metered? Equally, I would expect to see a difference in average water consumption between apartment dwellers and house dwellers. If there is a difference in the proportions of apartments in different regions, then this could be a potential independent variable. Otherwise the included independent variables are sensible.
Following your suggestion, we have now employed the variable household size in the set of potential explanatory variables (in fact, we had already used it, although this was not mentioned in the paper for the sake of brevity). As can be seen at the end of Table 3, the discriminating power of this variable is not very high, given that the average values for the clubs 1 and 2 are quite similar. In any event, we have also used this variable to improve the specification presented in Table 4, but the low values of the t-ratios associated to this variable led us to omit it from the final specification, as the results of Table 3 suggested.
- The paper has a nice appendix giving the data sources. I agree that the average price should include both the potable water service and the foul sewage service. Does the data source however include stormwater drainage, something which has no necessary linkage to the other two services? Again, on the appendix, does the source give any details of the method by which the figure of average water consumption was calculated or provided? The analysis hangs on these numbers being reliable and from a consistent methodology.
The methodology to obtain the data is described here:
https://www.ine.es/daco/daco42/ambiente/agua/metodolo.pdf
Please note that it is written in Spanish and we have not found an English version of this report.
According to this report, the variable that we have employed in the paper reflects the registered water consumption, which means “the volumes measured at users' meters (both community and individual)”. Consequently, it does not take into account stormwater drainage.
Additionally, the data are based on a poll, with the details of these polls reflected in this report, whilst this is the questionnaire:
https://www.ine.es/daco/daco42/ambiente/agua/cues_sum_agua.pdf
Again, we have not found an English version of this report.
We hope this can help to give a better understanding of the sources of the data employed in the paper.
- I would like to see some more discussion on the methodological problems that had to be addressed. The study is based on averages from aggregate data. This first raises the risk of the ecological fallacy, where differences between aggregates are interpreted to account for differences with each aggregate. A second problem is the problem of regression to the mean, we lose the variance within the population and equally lack knowledge of the original frequency distribution. A further issue in regard to the average price is that unless the price regime is one of a standard unit rate per unit volume, without any standing charge, then the average price per unit varies with quantity consumed. Quite often, the average price per unit shows a U-shaped distribution; a fixed standing charge on the one hand and a rising block charge on the other.
In our humble view, the methodology employed here does not suffer from the problems you mention.
First, we should take into account that our convergence analysis is based on the long-run trend component of the household water consumption. We think that this could at least alleviate the problems you mention.
Secondly, a particular case of the ones that you mention is Galton’s fallacy (related to the problems generated by the regression to the mean). This is so important in convergence analysis that it is the key to understanding why the standard b-convergence methods are not valid. In fact, the methodology employed here circumvents these problems by considering the evolution of the cross-sectional mean (sigma-convergence) and developing the method for testing whether this cross-sectional mean goes to 0 (which implies convergence).
With respect to the rest of the points you mention (first and third), we should note that our results are based on the relative position of a region with respect to the rest, then the relative behavior of the variables is more important than the absolute one, this being valid for both the PS and the logit methodology. This avoids (at least partially) the problems you mention, which of course are very important in other studies.
- Another country where there are quite marked differences in household water consumption between regions is Germany. Some years ago, I tried a cross-sectional analysis using various potential explanatory variables. I did not find the resulting multiple regression results very convincing partly because of the above reasons.
It is possible that the use of time series techniques, as is the case of our paper, could help to overcome this problem.
- As Valencia is the city I have visited most often in recent years, I was fascinated to see that it shows up in Group 1 with high per capita water consumption. A paper I found in consequence was: Maldonado-Devis et al 2021 “Panel Data Estimation of Domestic Water Demand with IRT Tariff Structure: The Case of the City o Valencia (Spain)”, Sustainability 13, 1414. This found significantly differences within the city region itself.
According to our data, the household water consumption in the Comunidad Valenciana region (which is composed of the provinces of Castellón, Valencia and Alicante) was 226 lid in 2018. This was the 2nd highest consumption, whilst the water consumption in Spain was 187 lid. Then, it is not so surprising from this point of view.
The results of the paper you mention are quite interesting, given that they suggest the existence of differences when more disaggregated data are employed, as we could expect. However, it is not possible to find data of this level of disaggregation in Spain as a whole (which is a pity, given that their analysis could offer a number of very good insights to better understand the evolution of water consumption in Spain).
- So whilst I would like to see some small editing changes, a very publishable paper.
We have edited the text with the help a native English teacher.