Author: Atsu Amegashie
On September 8, 2016, Dr. Mahamudu Bawumia, vice presidential candidate of the New Patriotic Party, delivered a lecture titled “The state of the Ghanaian economy – A foundation of concrete or straw”.
Dr. Bawumia gave a thought-provoking lecture. He raised valid and legitimate concerns about the NDC government’s fiscal management. Although he is sometimes partisan in his analysis, Dr. Bawumia has nevertheless set a precedent in our oral political discourse by relying heavily on socio-economic data in his periodic lectures on the state of the economy. His willingness to take on the task of comparative analyses of the evolution of economic performance during the tenure of the NPP and the NDC is refreshing.
However, I think Dr. Bawumia made a few questionable claims in his lecture. In this short piece, I examine one of his claims.
In his lecture, Dr. Bawumia claimed that:
“…. in dollar terms, under the NDC, the minimum wage declined from $2.12 to $2.02 by 2016 (i.e. by 4.6%). Between 2012 and 2016 (during the tenure of President Mahama) the minimum wage in dollar terms declined by 23.6%! In comparison, the NPP increased the minimum wage from the equivalent of $0.62 in 2000 to $2.12 in 2008 (i.e. by 244%).”: http://www.peacefmonline.com/pages/politics/politics/201609/291452.php
Commenting on Dr. Bawumia’s claim, IMANI observed that:
“In 2001, when the NPP took office, the minimum wage was 5500 old Ghana Cedis (dollar equivalent of $0.8). By end 2008 when their term came to an end, it was GHS 2.25 ($1.9). Using the US Feds Funds Rate and historical inflation rates to reset the value to 2001 constant rates indicate a real value of about $1.5. So Dr. Bawumia is not dead-on accurate, but his general point is valid. The NPP did roughly double minimum wages between 2001 and end-2008 in constant US dollar terms. The NDC, on the other hand, started with a rate of $2.1 in 2009 and is ending its term with a minimum daily wage of $1.9 (lower dollar inflation and treasury rates account for the lower discount factor). In effect, wages for the lowest earners in the economy have been stagnant under the NDC.”: http://www.myjoyonline.com/news/2016/September-9th/imani-alert-bawumias-speech-separating-partisan-politics-from-public-policy.php
What we should be interested is the real minimum wage: what a cedi can buy or the purchasing power of the minimum wage. It is very easy to show that if the nominal minimum wage (in cedis) is increasing over time but the percentage increase is smaller than the rate of depreciation of the cedi (against the dollar), the minimum wage (in dollars) will fall. But what does this tell us? What is the economic logic of using the dollar inflation rate to compute changes in real wages in Ghana, a country in which goods and services are predominantly priced in cedis? This method computes real wages or real minimum wages or computes some measure of real wages without using the inflation rate of the domestic prices of goods and services.
The only method of computing a real wage is to adjust the nominal wage by the rate of inflation or some index of the prices of goods and services consumed (i.e., a consumer price index) in a country. I call this the “standard methodology”. This is how real wages are computed in UK, Japan, USA, Germany, Singapore, etc. As an economist, Dr. Bawumia is definitely aware of this methodology. I do not doubt that for a second and do not wish to suggest that he does not know this methodology. Far from it. I have the utmost respect for his professional competence. However, Dr. Bawumia did not use the standard methodology to make his point that the NPP has a better record on the minimum wage. The point of this piece is to make the case that we should not depart from the standard methodology.
Dr. Bawumia may have used a nonconventional methodology because there is a relationship between exchange rates and inflation. But this is an indirect method based on a relationship between exchange rates and inflation that is not exact. As shown below, the standard methodology and the Bawumia-IMANI methodology need not give the same results. The nominal minimum wage (denominated in dollars) is not a measure of the real minimum wage, which should be our criterion in discussions about the minimum wage.
Even if the depreciation of the cedi affects the prices of imported items (i.e., what is known as “exchange-rate pass-through”), the consumer price index is used to compute the rate of inflation by using the market cedi-denominated prices of the goods and services, rates of inflation figures quoted by Dr. Bawumia in his lecture to support some of his claims. Instead of adjusting the nominal minimum wage (in cedis) using the domestic rate of inflation, the minimum wage (in dollars) uses the rate of inflation in the cedi-dollar exchange rate as the adjustment factor. In some cases, using the exchange rate (cedi-dollar) may give you answers that are close to those obtained when the CPI/inflation adjustment is used. But this occurs by chance. It is better to use the standard and correct methodology.
The analysis should be done by adjusting the nominal minimum wage by the domestic rate of inflation of the prices of goods and services. I have done this in appendix A below. It shows that between the time they assumed office and left office or July 2016 (in the case of the NDC), both the NPP and NDC (Atta Mills and Mahama) increased the real minimum wage. But there was a slight fall in the real minimum wage of about 3.7% during the most recent three-year period of president Mahama’s tenure (from July 2013 to July 2016). The computations also show that Atta Mills, during his 3.5 year tenure, increased the real minimum wage by 31.37% while Kuffuor increased it by 23% in his first 3.5 years.
Therefore, contrary to the results of Dr. Bawumia and IMANI, the real minimum wage did not decline under the NDC (Mahama and Atta Mills combined) between 2009 and 2016. The computations show there was an increase of about 20% in the real minimum wage from January 2009 to July 2016. However, the NPP increased the real minimum wage by a much bigger percentage (i.e., 200%), an impressive performance. In both cases (i.e., the NDC and NPP periods), it is important to note that this does not necessarily mean that the minimum wage was high enough to give the average Ghanaian worker a decent standard of living: absolute levels and changes in levels are two different things. Furthermore, our minimum wages are per day, not per hour.
The above results are subject to the caveat that it is better to look at the real minimum wage over various points in time rather than at only two points in time. For example, Kuffuor increased the real minimum by 23% in his first 3.5 years but increased it by 33.33% in his first three years. To make the point differently, suppose one government increased the nominal minimum wage by 2 cedis twice in a year (e.g., every six months) with the second increment occurring at the end of the year while another government increased it by 4 cedis once in a year and at the end at the end of the year? The increase in the real minimum wage between the beginning and end of the year will be the same. But to claim that the effect on welfare is the same will be misleading. Finally, the rate of unemployment (i.e., the probability of earning at least the minimum wage) is as important as the minimum wage.
By the way, is the minimum wage enforced in Ghana? Or is it only enforced in the public sector? Has anyone in the private sector ever been punished for paying an employee below the minimum wage? According to the US Department of State, in Ghana “… there was widespread violation of the minimum wage law in the formal sector…”: https://www.ecoi.net/file_upload/1788_1293377374_gha35914.pdf
Computation of the real minimum wage using the CPI
Sources of data: IMF, Ghana Statistical Service (GSS), AfricaPay, CitiFM
August 2005 IMF Country Report No. 05/286 Ghana: Statistical Appendix: Table 16: Ghana: National Consumer Price Index, 2000:Q1 – 2004:Q4:
GSS’ CPI Bulletins and Press Releases: http://www.statsghana.gov.gh/cpi.html
July 2016 minimum wage: http://citifmonline.com/2015/10/03/daily-minimum-wage-increased-to-gh%C2%A28/
December 2000: 5500 old cedis (0.5 Gh cedis);
December 2008: 2.25 Gh cedis
July 2016: 8 Gh cedis
Consumer Price Index (CPI)
December 2000: 182.9 (used last quarter of 2000 from IMF data)
December 2008: 271.46
July 2016: 180.3
To determine changes in the real minimum wage between December 2000 and December 2008, we need to find how much the nominal minimum wage of 0.5 Gh cedis in December 2000 was worth in December 2008. To do this, either (1) multiply 0.5 cedis by the CPI in December 2008 and then divide the result by the CPI in December 2000, or equivalently (2) multiply 0.5 cedis by (1 + rate of inflation between December 2000 and December 2008).
Using the data above and applying the formula in (1), we get:
0.5*271.46/182.9 = 0.74 Gh cedis.
This means that 0.5 cedis in December 2000 had the same value as 0.74 cedis in December 2008. Noting that the minimum wage in December 2008 was 2.25 Gh cedis, it follows that the NPP increased the real minimum wage by (2.25 – 0.74)/0.74 = 200%.
Given the availability of CPI data, the post-Rawlings NDC period, for the purpose of this analysis, is from January 2009 to July 2016. I note that from January 2000 to June 2013, the base year for computing the CPI was 2002. But in July 2013, the Ghana Statistical Service (GSS) changed the base year from 2002 to 2012. Therefore, the calculation for the NDC period will be based on using the CPI for December 2008 (base year: 2002) and July 2016 (base year: 2012), two different base years that are 10 years apart. This will not be comparing apples to apples. I deal with this problem in two ways:
(i) Statistical agencies publish “rebasing factors” that convert the values of an index from one base year to another. For example, see page 3 of this document by the US Bureau of Labor Statistics: http://www.bls.gov/cpi/cpimathfs.pdf
I could not find a “rebasing factor” at the GSS’ website. Fortunately, the GSS has published CPI numbers from January 2013 to November 2013 using the old base year 2002. For example, the values of the 2002-based CPI numbers from July 2013 to November 2013 are 459.4, 459.9, 450.2, 451.2, and 453.6 respectively. The corresponding values with 2012 as the base year are 113.6, 112.8, 112.0, 114.5, and 115.4 respectively. Comparing the values of the CPI shows that the values of the 2002-based CPIs are about 4 times the values of the 2012 based CPls.
Using a figure of 4 as an approximate “rebasing factor”, I multiply the July 2016 CPI value of 180.3 (base year = 2012) by 4 to get a July 2016 CPI value (base year 2002) of 4*180.3 = 721.2. Then we get
2.25*721.2/271.46 = 6.69 Gh cedis.
Therefore, 2.25 cedis in December 2008 was worth 6.69 Gh cedis in July 2016. Noting that the minimum wage in July 2016 was 8 Gh cedis, it follows that the NDC increased the real minimum wage by (8 – 6.69)/6.69 = 19.6%.
(ii) Suppose instead that we compute changes in the real minimum wage by looking at periods during the NDC’s tenure when the base year was either 2002 or 2012. Between January 2009 to July 2012 (the Atta Mills period), the base year was 2002. This 3.5 year period is shorter than the NPP’s eight-year period. The CPI in July 2012 was 412.4 and the (nominal) minimum wage was 4.48 Gh cedis. Then we get:
2.25*412.4/271.46 = 3.41 Gh cedis.
Therefore, during the tenure of Atta-Mills, the real minimum wage increased from 3.41 Gh cedis in January 2009 to 4.48 Gh cedis in July 2012, an increase of 31.37%.
Now consider the period from July 2013 to July 2016 (which has the same base year of 2012). The CPI in July 2013 was 113.6 and the minimum wage was 5.24 Gh cedis: https://www.ghanabusinessnews.com/2013/04/30/ghana-increases-daily-minimum-wage-by-17-for-2013/
Then 5.24 Gh cedis in July 2013 was worth 5.24*180.3/113.6 = 8.31 Gh cedis in July 2016. Therefore, there was a decline of 3.7% in the real minimum wage from 8.31 Gh cedis to 8 cedis during the most recent three-year period of Mahama’s tenure as president. Notwithstanding the approximation in the method in (i), it is difficult to argue that the reduction of 3.7% in the real minimum wage during Mahama’s tenure and the increase of 31.37% during Atta-Mills’ tenure would lead to the conclusion that the real minimum wage fell between the beginning and end of the Mills-Mahama tenure. The method in (i) showed an increase of about 20% in the real minimum wage during the post-Rawlings NDC period (i.e., Atta Mills and Mahama).
Using a shorter period for the NPP like we did for Mahama and Atta Mills, let’s consider the first three-and-half years of president’s Kuffuor tenure from January 2001 to December 2004. The CPI in the first quarter of 2001 was 197.3 and the minimum wage was 0.5 Gh cedis. The corresponding figures for December 2004 were 360.7 and 1.12 Gh cedis. Therefore, 0.5 Gh cedis in January 2001 was worth 0.5*360.7/197.3 = 0.91 Gh cedis in December 2004, an increase of 23% in the real minimum wage during the first 3.5 years of Kuffuor’s tenure.
The above analysis shows that between the time they assumed office and left office or now, both the NPP and NDC (Atta Mills and Mahama) increased the real minimum wage. But there has been a slight fall in the real minimum wage of about 3.7% during president Mahama’s tenure.
If we have figures on domestic inflation (and indeed we do), there is no reason why we should not use them. In computing real wages or real minimum wages, there is no justification for departing from the standard practice of adjusting nominal wages by the rates of inflation (CPIs) in the prices of domestically consumed goods and services. We should resist the temptation of using non-standard and less credible methods.
The above analysis is subject to the caveat that it is better to look at the real minimum wage over various points in time rather than at only two points in time. For example, suppose one government increased the nominal minimum wage by 2 cedis twice in a year (e.g., every six months) with the second increment occurring at the end of the year while another government increased it by 4 cedis once in a year and at the end at the end of the year? The increase in the real minimum wage between the beginning and end of the year will be the same. But to claim that the effect on welfare is the same will be misleading.
Let W be minimum wage in cedis.
Suppose 1 cedi is equal to D dollars. Then the minimum wage (in dollars) is $W*D (i.e., W times D). So, as claimed above, if D is falling at a faster rate (depreciation of the cedi) than the rate of increase in W, then W*D (the minimum wage in dollars) will fall.
We can rewrite $W*D as:
$W*D = $W/(1/D).
Note that, if as defined above, 1 cedi buys D dollars, then (1/D) is the number of cedis required to buy a dollar. So a fall in D (depreciation of the cedi) is equivalent to a rise in (1/D), which is a rise in the value of the dollar or what one might call an inflation in the value of the dollar relative to the cedi. Then given $W*D = $W/(1/D), it follows that the minimum wage (in dollars) uses the inflation rate in the value of the dollar to adjust the nominal minimum wage when the appropriate adjustment factor should be the domestic rate of inflation of the prices of goods and services.
Now consider a Ghanaian economy with the following features:
- The prices of all goods and services are quoted in dollars. Let this dollar price be P_d,
- P_d is fixed (i.e., the prices of items in terms of dollars are always constant),
- All payments are made in the cedi equivalent of the dollar price. Let the cedi equivalent be P_c.
Then P_c = P_d*(1/D).
I note that P_c will be an index of prices, not a single price. But for the ease of exposition, think of P_c as a single price. We may write the real minimum wage (in cedis) as:
W/P_c = W/P_d*(1/D).
Then given that P_d is fixed, changes in the real minimum wage will be driven by changes in only W and changes in the value, (1/D), of the dollar.
So given that the cedi is the main medium of payment in Ghana and subject to the caveat that P_c is not an index of prices, computing the minimum wage in dollars and using that to determine percentage CHANGES in the real minimum wage (or purchasing power of the minimum wage) makes sense if the conditions in (1), (2), and (3) hold. This may be referred to as full pass-through from exchange rates to domestic prices. However, at least one of the above conditions does not hold. In particular, the conditions (1) and (2) do not hold. And to reiterate, if we have figures on domestic inflation (and indeed we do), there is no reason why we should not use them to compute real wages (even if the three conditions above hold). In addition, as shown above, using the CPI gave different results.