Never Wanted to Be a Millennial

Access to justice: a brief geographical study in Singapore

Reading Time: 7 minutes

Introduction

(scroll down for interactive maps; you can click on the map markers!)

Since I’ve been learning programming on the side and on the job for a while now, and I’m taking a short break after my LL.M. to rest, reflect and learn some new (non-law things), I thought now would be a great time to formalise some of the things that I’ve learnt. Data vis excites me (the New York Times has amazing visualisations btw), and I’ve wondered how it would look like in advocacy. To those ends, I am just about to finish IBM’s Data Science Professional Certification. For my capstone project, to put into practice what I had learnt, I thought it would be interesting to examine data on the location of law firms in Singapore. Why would that be even remotely interesting? It turns out that it reasonably interesting. Besides the relevant observations, I also picked up on some trivial (in the sense of being the stuff of trivia nights) but nonetheless interesting observations. One of them is the prevalence of food & beverage venues throughout the island, regardless of whether the primary use of an area is commercial, residential, or industrial. Never thought I would observe the Singaporean love of food in this manner.

But I digress. Access to justice is an issue in countries around the globe. In Singapore, is it as much of an issue? Numbers can begin to tell the story, but I took a visual and geographical approach to tackling that question in Singapore. It focusses on brick and mortar law firms, rather than primarily digital ones, as they remain the go-to places for everyday legal matters. While I am an ardent proponent of LegalTech, and there is an ongoing shift towards remote working arrangements and the digital provision of legal services, brick-and-mortar firms are likely to be a mainstay of the legal profession for a while yet.

For simple private matters, are lawyers evenly distributed enough for people in residential areas in Singapore to have even access to lawyers[1]? My hypothesis is that most law firms would be concentrated in the Central Business District, in the area around the State Courts, and away from residential areas.

While this seems to be an obvious fact, it is not obvious that this is an ideal arrangement. For simple matters, factors like renown and reputation become less of an issue – as long as the lawyer is competent enough at handling the matter at hand. So, convenience and cost may be more relevant factors. Exploring this would be of interest to lawyers seeking new or cheaper places to site their practices, or for Singaporeans interested in knowing where to find lawyers closest to them, or more importantly, for increasing access to justice, especially in lower-income or high-density areas.

According to the Law Society of Singapore, there were 4834 registered locally-qualified lawyers and 1200 foreign lawyers in Singapore in 2017[2]. As of June 2019, about 5.7 million people live in Singapore.[3] Using these statistics, there is about 1 lawyer for every 944 persons (putting aside the fact that foreign lawyers would probably not be able to advise on issues of domestic law). In comparison, Uruguay has 1 lawyer for every 240 persons, while Israel has 1 lawyer for every 139 persons[4]. However, numbers like these do not portray the full picture. Other important questions are left out of the picture: are the lawyers good; do simpler matters take up a larger proportion of the caseload; do more commercially profitable matters take up a larger proportion of the caseload; which areas of law are most well-represented, etc.

For example, criminal law practice often sees fewer entrants and practitioners due to the lower margins of profit when compared to corporate law practice, for instance. This post focusses instead on the geographical spread of lawyers: whether lawyers are distributed evenly throughout the country, or close to population centres, where there is probably a greater need for community lawyering e.g. involve everyday legal matters like conveyancing, wills, family disputes, simpler disputes, simple contracts, etc. The current situation is probably due to law firms wanting to be closer to other law firms, the courts, dispute resolution facilities, other companies (especially for corporate law practices), and to harness agglomerative economies (costs savings that arise from clustering related businesses).

Data

The following datasets will be used:

  • Legal Services Regulatory Authority directory of law firms: to explore the locations of law firms;
  • Foursquare data on common venues: to explore actual land use zones or clusters;
  • Singapore Land Authority’s Singapore Planning Districts: for districting, clustering, and to explore planned use[5]; and
  • Singapore Land Authority’s onemap.sg data on the distribution of dwelling types in each district: to explore the various land use zones.

I will use the Singapore Planning Districts as a starting point. I will then use both Foursquare calls, as well as onemap.sg data, to find out where residential districts are. This will be done using visualisation and KMeans clustering (a machine learning algorithm that separates observations into clusters based on their proximity to a mean). I will then use the LSRA directory to examine where law practices are located, and their proximity to these districts.

Methodology

The methodology I adopted is as follows:

  1. Collect all planning districts from onemap.sg
  2. Calculate the centroids for each planning district
  3. Get the distribution of dwelling types in each planning district from onemap.sg: Housing Development Board Flats, Condominiums and Apartments, Landed Property, and Others.
  4. Get the most common venues in each planning district from Foursquare
  5. Carry out KMeans clustering on the most common venues to identify possible clusters (in other words, different zones of land use)
  6. Query law firm information from LSRA 
  7. Get the coordinates of their location from onemap.sg
  8. Query onemap.sg for which planning area that they are in
  9. Calculate number of law firms per district
  10. Show distributions on maps (this is the fun part!)
  11. Overlay law firm locations on
    • dwelling type distribution map
    • KMeans-clustered map

Results

As would have been guessed, law firms are predominantly situated in the CBD, with few having registered addresses outside the CBD. Few firms are actually situated in residential areas, or more densely-populated areas – at least as indicated by the number of dwellings.

The maps are discussed in the next section.

Brief Discussion

In terms of the LSRA’s data, some law firms did not have registered addresses (6 law firms), or had overseas registered addresses (5 law firms). Based on background knowledge, some law firms, like Hoh Law Coporation, might have multiple branches and therefore addresses. Based on a quick check of Hoh’s data on the directory, it seems that only the HQ is registered. So further work on this would have to look at the branches of the bigger law firms, if any.

Figure 1: Map of planning districts

Using the data provided by onemap.sg, boundaries around each planning district was drawn. Centroids (the ‘central point’ of each planning area) were calculated using the method outlined in the Jupyter notebook.

Fig 2: KMeans clusters (k=5, Top 10 most common venues, 200 venues within 3km radius of centroid)

While I did not take the additional step of testing which value of k was best, the 5 clusters observed seemed reasonable. The red clusters suggest transportation hubs at Singapore’s airport and port – along with Pulau Ubin and Sentosa. That seems anomalous, and would be interesting to explore. This was probably due to the clustering algorithm picking up on the presence of marine facilities. The light blue clusters are at Singapore’s CBD. The purple clusters are the most numerous, and suggests a merger of residential, commercial, and industrial areas into one type of cluster – ostensibly due to the prevalence of food & beverage venues in all of these areas (Singaporeans are known to frequent famous restaurants in otherwise industrial areas). If you can think of other possible reasons, do get in touch? The light orange area seems to indicate a military training area or heavy industrial area, based on background knowledge. The dark blue cluster seems to have picked up on the petroleum industry facilities on Jurong Island and Pulau Bukom.

Figure 3: Distribution of dwellings and locations of law firms

The following map highlights the distribution of dwellings, and pinpoints where law firms are located. From the map, it can be seen that law firms are primarily situated in the CBD, with sparser distribution in residential areas.

Figure 4: Distribution of lawyers

Using law firms’ registered addresses, this map highlights the number of lawyers in each planning district. From this, it seems that lawyers are primarily concentrated in one area within the Central Business District, and away from the densest residential districts observed in Figure 3. The gray areas indicate that no law firms are present in those areas. As hypothesised at the outset, this is probably due to being able to harness agglomerative economies, quicker communication with other law firms, proximity to the courts and other dispute resolution facilities, and finally, to other companies. The latter is especially important for corporate transactional and disputes work.

Figure 5: Map of K-clusters and law firm locations

This map contrasts law firms’ locations with the clusters picked up by the KMeans clustering algorithm. Law firms are primarily located in the light blue clusters, which coincide with the CBD, which again supports my hypothesis.

This brief exploration supports the hypothesis that law firms are primarily situated within the CBD, and away from residential areas. This is probably due to a variety of factors, as mentioned above. Beyond those factors mentioned above, it would be interesting to explore the proportions of revenue brought in by type of law practised. Presumably, the proportion of revenue brought in by corporate law practice would be much higher, and would also explain concentration in the CBD.

However, a more detailed analysis of this would perhaps yield more interesting insights into whether the advantages accruing from agglomerative efficiencies and convenient amenities are overstated. It might be more profitable to be situated closer to residential areas, whether because rent is cheaper, or because residents are more likely to patronise their neighbourhood legal practice for simpler matters rather than travel all the way to the CBD. In this way, it could be more economical to locate law firms in the manner of banks and their branches, rather than concentrating the entire company in one office in the CBD.

Pursuing this direction is perhaps even more viable with the shift to remote working arrangements given the ongoing COVID-19 crisis.

Future Directions

The following are possible steps which could perhaps yield further insight into this particular aspect of access to justice:

  1. Check for different values of k in the KMeans clustering algorithm
  2. Use the directory of lawyers
  3. Examine the proportion of legal industry revenue contributed by each practice area
  4. Examine the relationship between Point 2 and the specialisations of law firms (already present in the dataset)
  5. Examine income and population density metrics, if available
  6. Examine the extent of usage of digital legal services
  7. Examine, briefly, the relationship between actual land use and planned land use (this sounds like it would be a whole new rabbit hole)

Conclusion

Pursuing the future directions outlined above would arguably yield more insights into the question of access to justice, or perhaps prove irrelevant with the dawn of digital legal services. I am hoping for the latter!

If you are interested in reading the full report, please do get in touch!

Thank you to Stella Cao, Luke Wu, and Jerrold Soh for looking at drafts of this. All errors are mine.

[1] Assuming that access to lawyers means access to justice.

[2] https://learn.asialawnetwork.com/2017/08/30/how-many-lawyers-singapore-infographic/. The Legal Regulatory Services Authority seems to have 3991 lawyers on their database, as of 11 June 2020: https://eservices.mlaw.gov.sg/lsra/search-lawyer-or-law-firm/.

[3] https://www.singstat.gov.sg/modules/infographics/population

[4] https://abovethelaw.com/2018/02/the-most-lawyers-in-the-world/

[5] https://en.wikipedia.org/wiki/Planning_Areas_of_Singapore

hiyachai