Wednesday, January 14, 2015

How to buy votes when you can’t buy votes

Today, less than democratic regimes face a serious dilemma – how do you buy votes to win an election without becoming an international pariah. Unfortunately for a society and fortunately for an autocrat, the wheels of power and administrative resources an incumbent regime wields provide ample opportunity to manipulate electoral outcomes through what are otherwise legitimate activities related to state spending and coercion. In this blog post the term autocrat is used descriptively rather than in an evaluative or normative sense i.e. in a manner synonymous to a ruler of a less than democratic regime.

Notably, almost all governments in the world hold some form of election. Even fully authoritarian regimes such as North Korea  and Uzbekistan regularly hold elections after all. In hybrid regimes, which are neither fully authoritarian nor fully democratic (Armenia and Georgia among them), elections are often contested by multiple parties who see them as the legitimate means of gaining power. Political scientists have taken note and have come to call hybrid regimes which hold elections either electoral authoritarian or competitive authoritarian.

Elections in such regimes serve a number of purposes including legitimacy building on the international stage and enhancing stability at home. As Andreas Schedler has noted:

Electoral authoritarian regimes neither practice democracy nor resort regularly to naked repression. By organizing periodic elections they try to obtain at least a semblance of democratic legitimacy, hoping to satisfy external as well as internal actors. At the same time, by placing those elections under tight authoritarian controls they try to cement their continued hold on power. Their dream is to reap the fruits of electoral legitimacy without running the risks of democratic uncertainty. Balancing between electoral control and electoral credibility, they situate themselves in a nebulous zone of structural ambivalence.

This “nebulous zone of structural ambivalence” leads to elections which are neither clearly stolen, nor clearly free and democratic. Direct electoral fraud may damage a state and its leading bureaucrats’ reputations abroad, potentially lead to sanctions, civil unrest or even a revolution as occurred in Georgia in 2003. The presence of influential international organizations’ election monitoring missions increases the likelihood that the direct purchase of votes from the electorate, ballot box stuffing or other obvious violations of democratic norms will be publicized and punished in some form. Yet, the “democratic uncertainty” that accompanies fully free elections presents the possibility that an autocrat could lose their power. So, if an autocratic regime desires the benefits accompanying formal adherence to democratic rules, but simultaneously does not want to lose power or create instability domestically, it must find a means to illicitly secure votes without directly buying them.

Increasing social spending or creating employment, social benefit and similar programs is one way an autocrat can attempt to buy votes without directly buying them. Manipulating public spending in order to improve electoral outcomes was first termed a “political business cycle” by Nordhaus in 1975. His work hypothesized that politicians would favor pre-electoral policies that, in the short term, yield lower rates of unemployment and higher rates of inflation instead of policies that would encourage an optimal balance of both in the long term. In doing so, a government uses fiscal policy to improve the economic situation in the short term and, as a result, popular perception of a party’s performance immediately before elections to improve their chances at the polls.

While spending is one avenue for an autocrat to seek, coercion is a second option when spending alone may produce uncertain results. Notably, Bhasin and Gandhi, looking at presidential elections in every authoritarian country from 1990 to 2008, have argued that before elections “regimes will moderate their use of violence against ordinary citizens, while simultaneously directing state-sponsored repression towards opposition elites. Ordinary citizens are likely to experience greater repression after the election.” While direct, physical violence is one form of coercion autocrats sometimes rely on, a regime can also use fines and fees in order to punish the society it governs.

Autocrats thus may  have carrots (spending) and sticks (coercion) which they can use to manipulate electoral outcomes in their favor. Using the 2006-2013 state monthly income and spending data from the Georgian Ministry of Finance, this blog looks at social spending and other income, which consists of non-tax revenues from the sale of services and goods, and from fines and penalties. While the exact denomination and number of fines and penalties would be a better indicator, which would allow us to fully test whether Bhasin and Gandhi’s theory is accurate in the Georgian case, other income is used here as a proxy since the Georgian government has not published a full account of the fines it has administered.

In Georgia, a country which has consistently been ranked as having a less than democratic but not fully authoritarian regime by major democracy scoring indexes such as Freedom House and Polity IV, a clear political business cycle can be observed when looking at social spending between 2006 and 2013, the years for which monthly data on social spending and other income is available from the Ministry of Finance. The chart below shows, on the one hand, the monthly average of social spending for each election year, excluding the month before elections, and, on the other hand, social spending the month before elections. In every case, the latter has been higher than the former.

The clearest example of electorally motivated social spending in Georgia was in December 2007, in the lead up to the 2008 snap presidential elections that were held on 5 January, 2008. The dramatic increase in social spending makes particular sense when considering the options available to an autocrat to attract votes – spending and coercion. On November 7, 2007 the government violently cracked down on opposition protests in Tbilisi. These actions triggered widespread anger, and had the government continued to use coercion, it would likely have further decreased popular support for it. The incumbent regime realized it was left with spending alone – the “carrot” – to attract voters, and the precipitous rise in spending indicates this. While social spending remained stable from January to September 2007, varying from GEL 51.8 million to 54.8 million, it increased to Gel 66.4 million in October, to GEL 134.3 million in November, and to GEL 206.6 million in December.

In addition to attempting to ‘buy’ voter loyalty, an autocrat can also attempt to coerce it, be it physically or financially. Looking at the Georgian state budget’s other income line from 2006 to 2013, a fiscal coercion cycle appears in addition to the social spending cycle described above. In the chart below, pre-electoral highs in other income are shown together with the average monthly other income collected each election year, excluding the pre-electoral high for the given year. The pre-electoral high occurred two months before elections every election year except 2010, when it occurred one month before elections.

Note: The other income average (blue bars) includes the month before elections, except for in 2010, since, as noted above, in all other election years the pre-electoral high occurred two months before elections. 
Correction: This graph originally presented the pre-electoral high in other income as 140 million in 2006; this is the figure for other income for March, rather than August of 2006. The graph has been updated to reflect the correct average in other income and the correct pre-electoral high in other income. The pre-electoral income in 2008 was originally reported as nearly equal to the average other income. In fact it was significantly lower than the average, two months before the elections.

On all occasions except the 2008 January elections, the state collected a significant amount more in other income during pre-electoral highs. In 2006, the state collected 137% of the average monthly other income before elections, 186% before the May 2008 elections, 338% before the 2010 elections, 453% before the 2012 elections, and 171% before the 2013 elections.

While without detailed data on the denomination and number of fines and penalties it is impossible to prove that the Georgian government – UNM and GD alike – has been trying to punish the electorate after elections and elites before elections, as hypothesized by Bhasin and Gandhi, the regular dramatic increases in other income before elections suggests that the government is punishing elites before elections. This is well exemplified by the 2012 pre-electoral period, when the government fined Bidzina Ivanishvili and his Georgian Dream coalition a number of times in the lead up to elections.

Once again, the 2008 January elections appear to be an outlier. Other income dropped from GEL 41.2 million in October to GEL 10.5 million in November of 2007, the month that the snap presidential elections were announced. It seems reasonable to suppose that the government believed and feared that an attempt at further coercion, after putting down popular protests, would have been more costly than beneficial for their electoral outcomes – the government had exceeded their ’coercion budget’ so to speak. This likely explains the astronomical rise in social spending discussed above.

Autocrats can and do use carrots and sticks to manipulate electoral results in Georgia. This blog post looked at social spending as a “policy carrot” and other income as a “policy stick”. In the month before elections, social spending has consistently increased in Georgia, while two months before elections other income too has increased, with the only exception occurring in the aftermath of the events of November 2007. This suggests clear pre-electoral spending and coercive cycles in Georgia, implying that the government consistently attempted to manipulate electoral outcomes.

1 comment:

xcaucasus said...

I find the use of "autocrat" needlessly polarizing -- isn't this an issue for governments everywhere. Generally, I would hope you use super-neutral language when delving into such topics, for past, current and future governments.