Ghanaians go to the polls today to elect or re-elect a new Parliament and President. I have been analysing the survey data about people’s partisan preferences and motivations for voting in the public domain over the weekend.
The election is tighter than the researchers are predicting. Which is very interesting because almost all the serious researchers and pollsters are saying the ruling NPP is winning easily.
There is some evidence that turnout will be high among the base for both parties but there is a bigger net margin among independents who feel disappointed by the NPP’s decisions and actions in a number of major transactions. This group of voters may outnumber those who decide to just stay away out of disillusion about the political system as a whole. The data indicates that such trends will be most manifest in Greater Accra, which the opposition NDC is pushing very hard to flip. NPP’s hold over Central Region is, another bellwether swing state, on the other hand, seems much firmer. Which reinforces the split & tight verdict bubbling up from the models.
These developments have made what should have been a cake walk for NPP suddenly very competitive. Given the extent of battering the NDC received in the last elections, they shouldn’t even seriously be in the race this time as they chose to retain the same candidate.
The suspected apathy among the NPP base is, however, completely unfounded. NPP leaning voters have been energised by the intense NDC bombardment during the campaign, making for a fascinating showdown in this otherwise boring and drab election.
Another thing I found time over the weekend to look into is the role “corruption” will play in this elections.
Here, the argument that has been making the rounds is that “corruption” is not featuring at the top of voters’ concerns.
I have looked at the underlying sentiment data informing this academic analysis. I think the interpretation is suspect.
It is actually a well known issue in “social choice” mathematics. The likes of Gibbard, Satterthwaite, Debreu and Arrow have all commented on various aspects of this issue, and Condorcet’s methods have been applied for decades to handle some of the nuances.
When you ask people to pick one choice from a list of alternatives, you cannot proceed to establish an aggregate preference, which is how most of the researchers dismissing corruption as a factor appear to have done.
Let me illustrate.
Say there are ten issues of concern and you ask people to choose their top one.
12.5% choose issue A, and the rest of the votes are split equally among the remaining 9 alternatives. Of course the 12.5% of voters who chose issue A outstrip the other groups numerically but the interpretation is meaningless since issue B may be more widely represented as the second or third priority for all other respondents, gaining more weight as a result.
That is why the sound approach is to always have people rank their issues and then develop an ordinal scale to weight preferences/selections.
If you did that, you would find that whilst corruption may not be the number 1 or even number 2 issue for most voters, it may rank high enough across such a wide swathe of voters that it becomes much more important than simply using the percentage of people who chose it as their number one concern to measure its relevance.
There is also the sociological dimension. Corruption may not directly worry people except to the extent that it creates a sense of underlying unfairness, which then poisons the climate, leading people to demand more welfare from politicians.
So, the loud demands for bigger and better roads or free water and free education may actually emanate from a sense that a few people are enjoying the national patrimony and therefore that regardless of national debt or fiscal stability, all the money should go into visible indicators of welfare.
In that sense, resentment against corruption could manifest in other ways other than just a distinct focus or concern voiced out by voters.
If these inferences are plausible, then they can influence how voters evaluate a candidate’s policies and credit their performance, hence impacting their willingness to vote for or against them.
The challenge confronting this type of analysis in this election is of course that both candidates have not sufficiently addressed intense campaigns by their opponents now or in the past about their weak anti-corruption credentials so some “cancelling out” could have ensued.
With the online and non-scientific polls overwhelmingly favouring the Opposition and the scientific polls predominantly favouring the incumbent, my sense is that we should be cautious about the seeming consensus by professionals that the NPP is set for a smooth return to power.
My attempt to bridge the worlds of professional analysis and anecdotal evidence convinces me that the NPP has more to be apprehensive about than they think.
In the end, we hope that this elections shall turn out very well and that Ghana gets to add another enviable credential to its democratic trophy cabinet.
My colleagues at IMANI also can’t wait for the winner to be declared so that we can start putting their policies and political conduct under harsh scrutiny.
I am personally looking forward to chasing the Electoral Commission to account for its many procurement perversions whilst seeing off the current design of the present Government’s mineral resources monetization strategy.
May Ghana prevail!