design of democracy (3)
barriers to entry

We have tried to regulate away the positive feedback gain and have failed simply because it is impossible to decouple elected office from the accumulation of power. If the accumulation of power cannot be decoupled from elective office (and I discussed why this was impossible to do in my last post), then this accumulated power becomes a high barrier to entry for would be challengers.

Our democratic ideals say our candidates should be competing for votes based on the ideas they propose (their electoral platform) and their leadership track record (or the level of trust the public places on them). The reality though is that today's media soaked culture rewards high media exposure with high name recall -giving the advantage to candidates who get in the news often (be it political or entertainment news).

Incumbents have the advantage of a high profile job that gives them natural exposure in the news media. This is wholly part of the positive feedback gain. This media exposure along with access to social and political networks presents a high barrier to entry that would be challengers must overcome if they are to be on competitive footing against incumbents.

How do we compute the relative cost of this barrier? What toll does it put on our electoral system?

Investments and pre-investments

Running for an elective position requires resources. The higher the post, the more the resources are needed to get elected to the position. If you are running for president, you are competing for 30M-35M votes, and depending on the size of the field (number of candidates) will need to garner somewhere in the region of 12M to 17M votes. To get elected to the Senate, you will need to garner at least 12M votes. To get elected to the House of Representatives requires anywhere from 25,000 to 200,000 votes (depending on the district) .

You can monetize these voter numbers to see the cost of a campaign (i.e. P1=voter, then you will need P12M minimum to campaign for the senate). This is, at least in theory, relatively easy to do during the actual campaign period. We know of course that campaigning to get elected begins way before the official campaign period and so incumbent politicians who want to get re-elected will devote resources in advance of elections. This is where the comparative advantage - given by the positive feedback gain -comes in.

When the incumbent gets a CDF outlay for her district (and plasters her name all over the project), when she appears in public to give speeches or cut an inaugural ribbon, when she donates to community events, when she appears in the news, or when she glad hands local leaders -she is "pre-investing" resources in getting re-elected. She is also building name recall.

Likewise, a new candidate "invests" in brand creation and name recall when they elevate their public profile (via TV shows, or movies, or appearances in local fiestas, or being cited in the news) so as to get name recall.

Unlike funds spent on a campaign, the pre-campaign "investment" in name recall is nearly impossible to monetize. We need to find a way to measure the relative effect of these investments to see their effect on electoral contests.

Cost of voter acquisition: c as a function of v and r

We can attribute a per unit cost (c) to acquiring a voter (i.e. -getting a vote) by expressing it as the ratio of total votes needed (v) vs. total resources expended (r), such that:


The cost of acquisition "c", expressed as a per unit cost, allows us to compute the relative advantage of investing more vs. less resources in an electoral campaign. You can view it as the relative marginal cost of acquiring another vote or the relative "friction" of the system against a candidate. (The lower the number, the lower the marginal cost or the lesser the friction.)

We can use this relative unit "c" to grasp how high the barrier to entry is. It also allows us to illustrate the system dynamic.

Figure 1 (click image for larger view) shows that in a electoral contest over 50 million votes, investing 12M more units gives a candidate a relative advantage of -46c vs. a candidate expending 1M units. In other words, the marginal acquisition cost of another voter is 4c units for a candidate spending 13M, while it is 50c units for a candidate spending only 1M units.

The system dynamic means that (at least in large populations) more "r" gives you a lower "c."

An incumbent, with 4 to 6 years of media exposure along with access to resources and social and political resources, has easily invested huge amounts of "r" giving him a lower "c". The longer an incumbent stays in office, the more "r" is naturally invested in making "c" lower.

The system dynamic drives new candidates to expend a similar magnitude of resources to achieve a competitive "c". It also means that only persons who already have a high "r" have any chance of being considered as candidates. (We know this instinctively and we recognize "dark horse bets" and celebrate the unlikely success of "cinderella" candidates.)

The logic of getting celebrities to run for public office makes total sense given this system dynamic. Because celebrities come with built-in media exposure (read: pre-investment" -not to mention a strong "brand" based on their media typecast), a popular celebrity would have already expended high volumes of "r" which gives the celebrity a lower "c."

Next: Attempts to level the playing field by regulating "r"


the design of democracy (2)
the positive feedback gain

I've written about this before though we were basically dealing more with rate of replacement of the buffer stock (i.e.-getting new blood into the old congress) in that post.

Today, I will deal directly with the positive feedback gain and how it is impossible decouple this from the accumulation of power.

A few guideposts before we proceed.

  1. I use "positive" in a systemic sense, so I use the term in an amoral dimension. The feedback is "positive" only in that it encourages or amplifies the effect. (If were talking about sound then "positive feedback" would mean feedback that intensifies the volume.) I do not mean that imparts positive values.
  2. I will use the term "cost of each vote" (or "cost per unit" of votes) also without judgement. Cost can mean the actual price of buying a vote (corruption) or the cost of acquiring a voter, or the per capita cost of an electoral campaign (legitimate campaign costs) .
  3. Likewise, I use "resources" (which I will interchange with "power") in an amoral sense -it could mean money, or means, or social connections, or political networks or electoral machinery.

On the POSITIVE FEEDBACK GAIN of power and elections.

The feedback gain (which is also a "loop") works this way: getting elected allows a politician to accumulate power and "resources" (and by that I mean either political power, or wealth, or access to or control of actual physical resources, or access to or control of social networks or political organizations). These resources gives him/her or his/her scions a comparative advantage in the next elections.

The feedback is "positive" in that accumulating more resources gives the politician an even stronger comparative advantage. It is also a "loop" because, successive terms also amplify both the resources and the comparative advantage of the politician.

Attempts to decouple

We intrinsically understand that this accumulation of power, this positive feedback gain, is a threat to our democratic ideals. Democracy assumes a level playing field and we'd like to imagine that every candidate should stand for election (or re-election) on the merits of their worth as leaders. They should not gain undue advantage because their current position gave them access to resources and power that they can leverage in the next electoral contest.

(NOTE: I am personally ambivalent about re-electing incumbents. I think good leaders should get re-elected if they are to continue the good programs they have started.)

We have tried to decouple the positive feedback gain from the power of an elected official mostly by regulations or legislation. We have tried:
  • Imposing term limits -because we understand that the longer a politician stays in power, the more resources he accumulates.
  • Monitoring their wealth by requiring the yearly submission of Statements of Assets and Liabilities (SALs) -because it allows us to guard against the suspicious rapid accumulation of wealth. We regard wealth as a placeholder for an elected official's accumulated power.
  • Outlawing nepotism -to prevent incumbents from strengthening their hold on power via close-in connections in key positions. This is our way of trying to severe the inter-generational "loop" portion of the feedback gain. So, too, our next approach which is:
  • Proposing an anti-dynasty bill -to prevent incumbents from extending their hold on power by using their scions as their patsies and by leveraging their existing power to get their patsies elected. (NOTE: I think the congress will never pass this law as it is patently against their own self-interests.)
By and large, these attempts have failed because political power begets power and by default gives the elected official access to government and extra-government resources.

It is almost impossible to draw the line between the exercise of legitimate political power and the destructive leveraging of influence. Being in office alone gives the incumbent an advantage over any pretenders as the office gives the incumbent more visibility. (More visibility = higher name recall. Being elected gives you an advantage because it raises your public profile and makes you more familiar to the voters.)

We also have to admit that the power we bestow on elected officials is inherent in the job. That's why we elect them in the first place, to give them the political power to produce results for the electorate. They need the power to get things done. The last thing we want is an impotent office and yet it is very difficult to draw a strict line (although we have hundreds of lines in our legal books that attempt to do this) between the execution of office and the leveraging resources to get re-elected.
Example: When an incumbent senator visits a province to consult with the citizens in aid of legislation -does this opportunity for glad handing not give him a comparative advantage the next time those citizens come to the ballot?
The best we have come up with is a rule that prevents the use of public resources 90 days before an election. But, like the example above, campaigning to get re-elected can come in many subtle (read: totally legitimate) forms, even before the campaign period actually starts.

(In its most basic form, this is what the pork barrel is all about. The CDF and special insertions in the budget, allow the incumbents to "invest in their brand" and sell themselves to the electorate ahead of the elections. Hence the preponderance of large billboards attributing credit to the proponents of CDF funded projects. --Building political networks are also part of this early investment of resources.)

Clearly, our attempts to legislate a decoupling of the elected office from power as an advantage, have been largely ineffectual. The positive feedback gain is so inherent in the power of an elected office that incumbents are more likely to get re-elected, no matter how terrible their legislative (or executive) record. Their children are also more likely to be elected to their position -or to other political position. (Past electoral counts will bear this out and this is true not only in the Philippines but in almost all representative democracies.)

Apart from giving incumbents an undue advantage, the positive feedback gain has also raised the barriers to entry to elected positions. New entrants to the political game (no matter how able their leadership track record) must either bring in comparative resources vs their opponents or invest ahead in name recall so as to gain a comparable competitive position to the incumbent. (Which is why celebrities are such good fodder for candidates -they come built in with popularity -i.e., they are already a "brand" that the voters recognize and can stand their ground in the battle for name recall.)

We have also tried to regulate away this increasing costly barrier to entry (a side effect of the positive feedback gain) as another attempt to level the political playing field.

NEXT UP: Barriers to Entry

Image credit to ulibrskr's flickr photostream.


the design of democracy

The questions you ask limit the answers you can come up with. -Michele Tepper | FrogDesign
A lot of time has been spent (and probably will be spent) on debating the merits of a unicameral vs. a bicameral legislature, of a parliamentary vs. presidential system.

I know the legal luminaries, the experts on public governance and political science have weighed in on the project along with a thousand pundits and bloggers so I've avoided putting in a comment.

What I haven't seen though is the problem tackled by a design mindset. (See the UK Design Council's RED project, as a reference.)

The designer looks at both users and end results and then then traces the dynamic back to driving forces that move the users and shape the results. The designer looks at the dynamic not the rationales and rethinks the dynamic to leverage the driving forces. The designer begins by asking the right questions and understanding how the system operates (in its current form) before positing possible solutions. The design approach is agnostic and amoral in its analysis while producing ethical and forthright solutions.

So here goes. I will take a break from mind-mapping our city. Over the next few posts I will tackle the question of the design of democracy in our country.

Parliamentary or presidential? Unicameral or bicameral? I think the debates miss the problem entirely. Whatever form of government we choose (or not choose) the essential problem we face is corruption at the polls -and the successive doubt and distrust of both the results and the government-elect. No subsequent Philippine state will achieve legitimacy until we remove the cloud of doubt that hangs over our electoral process.

The doubt is built on two factors. One is the suspicion of corruption, of rigging the polls and wholesale vote buying.

The other is the seeming surfeit dearth of worthy candidates from which to choose our leaders. As Willy Priles opined in one of his recent posts, we seem to be forced to select from among the dregs that the parties serve up for national posts. The two major flavors the parties seem to be intent on foisting are ex-media personalities or scions of established politicians. Both trade on name recognition. The former from extensive media exposure, the latter from news familiarity and name recall.

You can change the form of state any which way you want but until we address this two-pronged doubt, all successive governments will be (in the best case scenario) treated with benign distrust or assailed with questions of legitimacy.

The quick answer is to reform the comelec. Build integrity into the agency charged with safeguarding the elections and people will trust electoral results. That's why we've been trying to automate the voting booths. But computerization and a trustworthy Comelec is a partial answer. Computerized systems can also be rigged and hacked. And even the most highly regarded officials can have their integrity swiftboated. All a spinmeister need do is to create doubt to start the public sliding along the slippery slope to distrust, while the media feed the conflagration by playing on the hesaid/shesaid conflict.

Evenso, computerization and a total overhaul of the comelec will not even begin to tackle the question of serving up and selecting would-be leaders nor will it severe the dynamic that warps it from its roots in nepotism and media manipulation.

The solution must go deeper. We have to look at the dynamics driving the electoral problems and look at the processes driving the selection of candidates.

I see three driving forces that are interconnected:
  1. The strong positive feedback loop -power begets power. power is a barrier to entry.
  2. The power of filters and information agglomerators - modern popular elections have 50M voters selecting leaders they will probably never meet in person so the game belongs to the marketers and image keepers while campaigns are not about governance but about attacking and defending brands.
  3. The relative cost per vote in very large elections - which makes votes more expensive by bulk but cheaper by the unit. (Here, I use "cost" in an amoral sense.) e.g., When selecting a mayor, each vote can be as cheap as 1/1,500,000 per unit (which is the ratio of 1 vote vs. the total number of would be voters). Selecting a senator or a president prices the unit at 1/50,000,000.
I will tackle each of these driving forces in successive posts and prototype some possible design approaches to mitigate the effects or even find solutions.

Image from of twotone streetart's flickr collection.

Quick Links

Notable posts on the metro