Hip Hop and the LAPD

I was recently asked whether I had any data to support the claim that the number of mentions of the LAPD (Los Angeles Police Department) in hip hop lyrics has fallen off in recent years, after peaking in the mid-1990s. The LAPD attracted a lot of angry attention from hip hop artists in the wake of the 1992 LA riots, triggered when four LAPD officers were acquitted of using excessive force in the arrest and beating of Rodney King

On the face of it, this claim should be easy to check: use a lyrics database to search for hip hop songs mentioning “LAPD”, find when they were released, and look at the trend. In practice, it turns out to be rather more complicated.

There are several online lyrics databases, which vary in coverage and useability. Most promising were lyrics.com and genius.com, which can be searched for lyrics containing “LAPD”.1 Genius.com also returned terms such as “lapdance”, whereas lyrics.com did not, so I chose the latter.2 Although lyrics.com allows you to restrict the search to hip hop, it appears that most songs are uncategorised, so this resulted in just six songs, compared to 244 from an unrestricted search (the vast majority of which are clearly hip hop).

From the lyrics.com list I obtained the song title, artist, year of release, a short biography of the artist, and the lyrics (or, in a few cases, a short extract).

Plotting the number of songs by year of release produces the following chart, showing a peak in mentions of LAPD in the 1990s, as expected, and a significant fall during the subsequent two decades…3

Job done! Alas no. Despite showing what we would like it to show, there is a major problem with this chart. There are no songs after 2020, and the total number is clearly much lower than the expected 244 songs. This is because lyrics.com included the year of release for only 75 songs – less than a third.

Manually checking the same songs on genius.com, which has better data on release years, brought the total number of known release years up to 208. I also discovered a few duplicates in the data, reducing the total number of songs to 226.

Now with years for 92% of the songs, the chart can be redrawn…

We now have plenty of songs from the 2020s, but there is a huge increase in mentions of LAPD in the last few years. What is going on? After quite a lot of trial and error, I identified two issues: the prevalence of cover versions, and changes in the population of hip hop artists.

Cover versions

There were several duplicated lyrics in the dataset, sometimes by different artists, and in searching for release dates I also found examples where the correct song title was attributed to a different artist. Further research showed that cover versions are surprisingly common in hip hop: out of 93,000 hip hop tracks listed on Musicbrainz, for example, almost a third have the same title as an earlier track,4 suggesting that many are versions of the same song by different artists.

This increases the potential for duplicates, and therefore double counting in the data. Perhaps more significantly, songs are often credited to the performers of cover versions, rather than to the original artists, meaning that release dates were sometimes much later than those of the original tracks.

So the apparently large number of LAPD songs in the last ten years is (at least to some extent) due to double counting of cover versions (i.e. the same song by different artists), and release dates that are later than those of the original songs.

To find the original release dates I searched for the LAPD song titles on Musicbrainz, and took the earliest release year for each. This is not 100% reliable, as not all songs are listed on Musicbrainz, and there are several examples of different songs having the same title, and of versions of the same song having different titles. Some hip hop artists, for example, like to exchange an “s” for a “z”, so a song title including “boys” might have it as “boyz” when recorded by a different artist.

Removing the duplicates and adjusting the release years to those of the original track (as far as possible) changes the chart to the following…

There is still a large increase in the number of mentions of “LAPD” over the last ten years, but it is less extreme than the previous version.

We could decide to keep the cover versions in the data, as performing an old song mentioning LAPD is (almost) as much of an artistic choice as writing those lyrics in the first place. Having said that, tracking down all cover versions of a song is probably at least as difficult as eliminating duplicates. I suspect there is also a grey area in hip hop, where a cover version may include some lyrics taken verbatim from the original, plus other sections that are modified, and perhaps some new material as well. Whichever approach is taken to include or exclude cover versions, a degree of uncertainty and error is likely to remain.

Hip hop population changes

In the 1990s, musicians needed a professional studio to record their songs, which were manufactured and distributed as physical CDs (or perhaps on vinyl or cassette). Listeners would probably encounter new music by hearing it on the radio, or perhaps on TV, in a concert or club, or via friends. Today, in the 2020s, anybody can make a high quality recording at home and promote it via streaming or digital download services without the need for a physical copy. Increasingly, listeners are discovering new music via personalised recommendations and playlists from sites such as Spotify and Apple Music.

Consequently, in the last few years listeners have had access to a rapidly expanding pool of artists and their songs. Despite that growth, the market (in terms, for example, of overall airtime or financial success) remains dominated by a relatively small number of well-known established artists – the vast majority of new artists and songs have very few fans or listeners.5 To try to quantify the effect of this, I gathered some extra data – from Spotify, Wikipedia and Bandcamp.

Spotify listeners

The music streaming service Spotify gives, for each artist, their total number of monthly listeners – a number that ranges from zero to over 100 million (at the time of writing Billie Eilish is at over 107 million).

Redrawing the previous chart, split by the original artists’ number of Spotify listeners, gives the following…

The coloured segments represent the number of Spotify listeners, ranging from 0-9 for the darkest red band (top) to 10 million or more for the dark blue band (bottom). There are few mentions of LAPD by artists with under 100,000 Spotify listeners until 2015, and almost all of the growth in mentions of LAPD in the last five years has come from artists with fewer than 1,000 listeners.

Of course the current Spotify score is not a perfect indicator of fame or obscurity for historical data, as artists rise and fall in popularity over time. Ideally, we might use the artists’ numbers of listeners at the time the songs were released, but that data is, as far as I know, not readily available.6 Nor is there an obvious alternative source that might include such historical data: the record sales charts (Billboard, etc) would get round this problem, but would only cover the very big names, so would be much less useful. In fact, as far as I can tell, none of the songs mentioning LAPD ever seems to have made the Billboard hot 100 singles chart. Most were included on albums but not released as singles.

Wikipedia pages

Somewhere on the route from obscurity to fame there is a point at which an artist acquires a Wikipedia page. Helpfully, there is a Wikipedia page listing hip hop artists,7 and a simple-ish way of finding the original creation date of any Wikipedia page.8 The distribution of Wikipedia page creation dates for the 5,704 currently listed hip hop artists and groups looks like this…

The peak around 2005 says more about Wikipedia than about hip hop artists, as the site (launched in January 2001) was still growing and being populated with articles. This had settled down by around 2010, after which there was a relatively steady rate of 200-300 new hip hop artist pages per year. There appears to be a gentle upwards trend from 2010 to 2020, but a slight decline over the last few years. Hip hop, as a genre, may well have reached a point where the number of artists reaching the top levels of fame has peaked and is starting to decline.

Bandcamp

If the population of established hip hop artists is growing more slowly than before, the same cannot be said of the other end of the market. As the LAPD data indicated, there are many new artists trying to get onto the bottom rungs of the ladder of fame. Bandcamp is a popular site for new artists (and some more established ones) to promote their music via streaming or download, and it tells a very different story from the Wikipedia chart above.

Bandcamp has a useful “surprise me” function which generates a list of 60 random hip hop tracks. Each track links to an album page, which gives (among other details) the release date. Using these features, I was able to generate the following chart…9

Clearly new hip hop music is thriving, and growing, on Bandcamp!

The sampling process allows us to estimate the total population of hip hop albums and artists. The random sample of 5,937 included 5,831 different albums: 5,726 were sampled once, 104 appeared twice, and one occurred three times. The total number of albums can be estimated as (size of sample) * (number of distinct albums in sample) / (number of repeated albums in sample) = 5,937 * 5,831 / (104 + 2 * 1) = 326,591. A similar calculation produces an estimate of about 52,000 for the population of hip hop artists.10

So there are perhaps a third of a million hip hop albums on Bandcamp, by over 50,000 artists, mainly from the last ten years. That is a HUGE number! For comparison, as we have seen, Musicbrainz (a reasonably comprehensive catalogue of physical recordings) lists about 93,000 hip hop tracks (not albums), and Wikipedia has pages for 5,700 hip hop artists. It is easy to see how even a tiny proportion of minor artists getting credited with mentions of LAPD on a lyrics database can easily distort the apparent trends quite significantly.

How is LAPD mentioned?

So far we have simply counted songs and not looked at the lyrics themselves. Can they tell us any more?

The short answer is “maybe”, but it is difficult to identify meaningful trends in the way that LAPD has been portrayed in hip hop lyrics, for three main reasons…

  • It is quite a small sample, once you break the songs down by period;
  • There is a lot of variety in how LAPD is referred to, so the differences within any period are much larger than those between periods (to use an engineering analogy, there is a very low signal to noise ratio);
  • The language of hip hop can be very difficult to parse. There are many abbreviations, slang terms and swearwords, and the lyrics are often composed of quick-fire short, sharp ideas and phrases which make any underlying narrative or sentiment hard to determine from the words alone. Reading the words on a page is no substitute for hearing them performed – much of the meaning comes from the rhythm, rhyme and articulation that are characteristic of the genre.

Hip hop and its lyrics have been widely studied, and there are experts out there who could probably say more than I can about the context in which LAPD has been mentioned. I can only offer three observations, from a brief examination of the lyrics in the dataset.

The first is that, as you would expect, there seems to have been a shift around the time of the 1992 riots. Before that, LAPD was largely portrayed as part of society, getting on with doing what you would expect a police service to do – good or bad, depending on your perspective. After the riots, the references become almost entirely angry and critical, with suggestions that LAPD is racially biased, unfair and heavy-handed. These sometimes refer to specific incidents, or may be more generalised.

The second observation is that LAPD has become an emblematic target for general anti-police or anti-authority sentiment, not just in Los Angeles but more widely. This is best illustrated by the extent to which mentions of LAPD appear in other countries and languages: the dataset contains songs in German, Polish, Swedish, Czech, French, Italian, Punjabi, Russian and Greek. The following chart shows how LAPD has been referenced by artists from the USA and elsewhere, with most of the international spread taking place over the last five years.

The third observation is simply that there are no lyrics in the dataset that are positive about LAPD (although a few are neutral). This is unsurprising, as the “Gangsta Rap” form of hip hop – the style of most of the LAPD songs – is generally critical of authority. I doubt there is a police force anywhere in the world that is regularly praised by hip hop artists!

Conclusions

To return to the original question, based on the above analysis, we can say that there is evidence to support the qualified claim that references to “LAPD” have declined since the late 1990s in original hip hop tracks by established artists.11

However, LAPD continues to get plenty of mentions from the rapidly expanding world-wide population of new and little-known hip hop artists. Many of these are cover versions (or perhaps borrow at least some lyrics from previous songs), and the LAPD is often simply a convenient emblematic target for general anti-police sentiment.

This investigation has encountered a complex and changing set of data, and it is difficult to make sense of it and to work out exactly what is going on. The lesser-known end of the hip hop market is so large and varied, difficult to quantify yet easy to access, that – with new music increasingly being discovered via personalised recommendation algorithms – it is possible that no two listeners will experience trends in new music in exactly the same way. A analysis of the numbers, imperfect as it is, is perhaps the only way to get an objective view of what is becoming an increasingly subjective and personalised musical world.

Cite this article as: Gustar, A.J. 'Hip Hop and the LAPD' in Statistics in Historical Musicology, 30th October 2024, https://musichistorystats.com/hip-hop-and-the-lapd/.
  1. Both also picked up “L.A.P.D.” automatically.
  2. The two sites have considerable overlap, so there is little benefit in using both of them.
  3. The five year bands begin at the start of the labelled years, so the first bar, for example, includes releases during 1980-1984, the second 1985-1989, etc.
  4. Ignoring differences in capitalisation, spacing and punctuation.
  5. This is always the case in any genre. See my previous publications on this topic – such as this article.
  6. In any case, Spotify did not start until 2008.
  7. There is another for hip hop groups – I combined the two.
  8. For details, see here.
  9. The sample was 6,000 but I discarded a few links that were for merchandise rather than albums.
  10. These estimates are approximate as they are based on a random sample, but they also use an approximate formula. If you are really interested in the theory and details, see this page. The calculation also assumes that Bandcamp’s “surprise me” function gives each album an equal chance of being selected. There will be some distortion due to different numbers of tracks per album, but otherwise I suspect that this assumption is probably reasonable.
  11. The discussion above illustrates the difficulty of being precise about the meanings of “original” and “established” in this statement.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.