We’ll continue to consider irregularities of Moldova 2019 election. Bow we’ll turn to the constituencies in the rural area. We’ll begin our study from several general notes and then we’ll define what object will be the study case.

The key problem is that, in contrast to Chisinau and Balti where all the constituencies are assumed regular by default, as there’s no evidence to the contrary, some of the other constituencies don’t match the normal distribution. Such constituencies are listed below:


  1. Three extraterritorial constituencies inherently are significantly different from all the other ones by the income level, gender and age distribution, etc. In addition, people outside Moldova has another level and cluster of main information about Moldovan political topics and political life than residents.
  2. Two constituencies in Transnistria are irregular at least because the voting was conducted outside the enclave on the Left Bank. I.e. the problem of provision of transportation to the polling stations was crucial for the Left Bank residents, causing severe irregularities. Many Moldovan politicians and activists paid attention to it. We’ll add that the voting on the Left Bank is historically different that on the Right Bank. It can be easily seen from the voting results in this part of the rural area of the Dubăsari district, which is under control of RM. In this case, there is a difference in principle, and so everything related to the voting results in the Left Bank will be excluded from our consideration. This should be a subject of a separate study.
  3. Two constituencies in the UAT Gagauzia and one in the Taraclia district. These three constituencies have ethnic configuration completely unlike that of the rest of Moldova, hence having completely different political configuration. So, Gagauzia and Taraclia also should be considered separately.


Thus, our study is limited to 30 constituencies (excluding some rural communities of the Left Bank belonging to the constituency c19 (Ivancea district)). We’ll discuss the existing statistics to assess its credibility and regularity and then we’ll consider the situation in the selected districts and constituencies explicitly.

We will notice that (as in the previous chapter) the statistics is “uneven” which in itself indicates the presence of irregularities. One can suggest that the political situation could have completely changed for 4.5 years, so the statistics could have changed too. However, as it can be seen in the example of Chișinău and Bălți, the main results haven’t significantly changed, but some indicators have changed a lot. This clearly shows the presence of irregularities and electoral fraud.

For better visibility, we will start from the middle and not from the beginning.


Villages by ethnic 2014 2019
Predominantly ethnic Moldovan 77.5 77.3
Mixed 15.4 15.4
Predominantly ethnic non-Moldovans 7.1 7.3

Table 7. Share of village elector vote by ethnic component 2014-2019.


We get just an ideal alignment preliminary very difficult to suppose. The indexes from the table 7 are very important one:

  • ethnic component in the rural area is changing very slowly,
  • turnout at 2019 looks like absolutely the same compare with 2014,
  • herewith, number of valid votes in the rural segment have significantly reduced – from 792 thousands to 697.


Therefore, indexes from the Table 7 could be used like an evidence that “elections are clear” and “statistics is quite normal”.  However, if we’ll restart from beginning, further alignments will not be so perfect


Moldova by area 2019/2014

(absolute figures), %

2014 2019
Rural 87.9 54.5 55.2
Urban 85.7 45.5 44.8
Incl. Chisinau and Balti (with suburbs) 82.5 25.4 24.1

Table 8. Share of Moldova elector vote by urban/rural area component 2014-2019.


Totals from Table 8 obviously contradicts with well-known practical situation in Moldova.
Chisinau and Balti are less suffering from the washing out of population than rural zone. Moreover, there is significant migration from Moldovan rural zone to Chisinau and Balti. There is impossible that in such situation share of rural votes will increase.

We’ll take indexes from Chisinau and Balti from the previous parts of the project like a base with account of irregularities. In this case, ratio from 2019 to 2014 in the rural zone couldn’t be more than 75-76% without falsifications, but we get just 88%.

We can preliminary suppose that ratio of irregularities could be defined at 12-15% and this is the index of the very high level of fraud. Further, we will try to prove these indexes. There were no signs of such irregularities in Moldova never before.

Particularly noteworthy these indexes of Chisinau and Balti (most wealthy areas in Moldova) significantly fell down compare with the centers of districts. Therefore, it would be interesting to continue decomposition of the most important electoral factors.

Moldova by urban area 2019/2014

(absolute figures), %

2014 2019
Chisinau 82.6 47.2 45.5
Suburbs of  Chisinau 86.6 10.3 10.4
Balti with suburbs 81.6 8.8 8.3
Centers of districts and small towns 90.8 33.8 35.8

Table 9. Share of Moldova elector vote by urban area component 2014-2019.

Obviously, this alignment again is impossible. Most wealthy Chisinau and Balti significantly fall down, centers of districts become the principal driver, Chisinau suburbs with quite normal demographical situation have no changes at all


  1. Further decomposition shows that the high indexes at some rural area and centers of the districts took place only in several districts and SMC. DPM and Sor party won all this constituencies. Therefore, the high turnout again is tightly correlated (as in Chisinau was) with extra votes for DPM and Sor. This correlation automatically is the sign of irregularities and falsifications.
  2. Statistical indexes become less biased only in Chisinau, Balti and some centers of the districts free from DPM power control. All those municipalities have less irregularities and falsifications than it has been in rural zone.
  3. Obviously, all election fraud begin from the centers of districts, just there were located DPM “smart groups”, but often it was more difficult to falsify at towns than in villages. Therefore, fraud has begun in centers of districts, and then went at rural zone. It wasn’t very “successful” at towns, but was necessary to launch “process of fraud” from the “smart center” and it was the essential “element” of falsification plans.


We restart with totals analysis from turnout in the Table 10. There are swings between 2014 and 2019 by all main geographical, ethnic and kind of municipality categories.

Moldova by different categories Right-wing Left-wing DPM
2014 2019 2014 2019 2014 2019
Predominantly ethnic Moldovan 37.9 28.0 38.7 33.8 22.3 38.2
Mixed 26.2 16.7 57.1 54.3 15.7 27.9
Predominantly ethnic non-Moldovans 9.4 2.3 82.5 83.1 7.5 14.6
Chisinau 45.4 40.3 45.1 48.5 9.2 11.2
Suburbs of  Chisinau 55.7 48.1 29.4 29.1 14.4 22.8
Balti with suburbs 20.9 15.6 67.9 73.7 10.6 10.7
Centers of districts and small towns 32.3 26.2 50.2 54.8 15.6 19.0

Table 10. Results of different political segments within different electoral categories 2014-2019.

Totals again are in contradiction with statistics. Therefore, we may make some intermediate conclusions before we will study cases of districts and single-member constituencies (SMC).

  1. DPM has maximal growth of support only in the two main categories – predominantly ethnic Moldovan villages and ethnic mixed villages. Nevertheless, DPM is not agrarians, not “party of village”. Moreover, there are no any comprehensive parts in the DPM program, which would be devoted to the agricultural or agro-industrial breakthrough. It would be just impossible, regarding to the social-economic situation in Moldovan rural zone.
  2. Such alignment is strictly impossible by the viewpoint of electoral campaign sociology and electoral process at whole. All campaigns in the area are going through the center of district, through any urban agglomerations, where are located the main district media, live political activists and local counsellors, who are the base of local branches of political parties and “smart groups”. So any growth of votes at villages has to be repeated in the district centers (not obviously the same but substantial), but in our study case it isn’t.
  3. At least. In must be the substantial growth both in the villages with predominantly ethnic moldovan and predominantly ethnic non-moldovan component, if it would be the message from any “agrarian” party, but also it wasn’t. Therefore, we see irregular alignment, where all DPM “growth” takes place out of the main statistical, sociological and political patterns.
  4. We just begin from the centers of districts trying to explain. Turnout in the centers was the most exaggerated one and evidently irregular. These indexes show that the centers of districts predominantly were and the “centers of falsifications” in the most of area constituencies. At the same time very small “growth” of DPM support in the district capitals shows that the ratings of DPM even in small towns were such low, that it was impossible for DPM to “make huge fraud”. At first the reason was, that in the towns there was more confident monitoring over the process of voting, other parties’ activists prevented DPM from the more numerous falsifications. Therefore, lot of forgers had to refuse from their mission because they met active counteraction from their rivals and were not able to continue fraud attempts.
  5.  However, it was the mostly different situation in the ethnic moldovan and mixed villages. We must remember the 2015-2016th situation (local elections, then the DPM government formation). Before Vladimir Filat arrest, the power control in districts was shared between LDPM and DPM. After Filat’s arrest LDPM have collapsed very quickly and all this “power control” was immediately captured by DPM together with the presidents of the districts and with the mayors of communities who were forced to enter DPM. PAS of Maia Sandu that took some part of “LDPM politic heritage” didn’t take any “power control” in the districts, PSRM had not such possibilities at whole. Therefore, the DPM falsifications in the villages mostly were not stopped, because there were not enough “counteractivists” from other parties, mayor were the part of DPM and mainly took part in the fraud.
  6. In the villages with non-moldovan ethnic component the situation is rather different. There are enough activists and it was enough control from the part of PSRM. All polling stations were created with the participation of PSRM local branches and mostly it was enough to escape fraud. Therefore, the size of falsifications was the minimal one and did not influenced at the main indexes in this category.
  7. The main. The character of fraud, uneven and spoiled statistics, sometimes quite normal, sometimes impossible, tells us that fraud was prepared in depth. Sometimes normal statistics is the result of adjustment of voting protocols to the needful totals, the results of some “preliminary job” with totals, which “called” some hours of disconnected server during the process of vote’s calculation.


DPM was not able to switch the totals in their favor more significantly, but they gained just good totals, which are based on fraud and mass falsifications.

Tricks in districts and constituencies.


We’ll begin case study about districts and constituencies and will try to determine the formula of DPM “success” in areas which made them winners in the most of them. At the first, we’ll consider the districts and the constituencies in the districts. It gives us possibilities to compare both totals from 2014 and 2019 parliamentary elections and 2019 vote by the different electoral systems. Preliminary we must exclude the districts that were not normal just from distribution of the votes.

Districts Nisporeni and Orhei have their own personal component, connected with Vladimir Plahotniuc (Nisporeni) and Ilan Sor (Orhei). Specifics of those districts are clear by totals and well known by facts. So they would be excluded from case study. Then we’ll exclude some “incomplete” districts, cut down not to form their “own” constituency in the boundaries of this district. We’ll study them in the next paragraph, but they are not normal in the means of district – constituency relations. The result deviation in the incomplete districts is not obvious for deviation in the constituencies, so statistics for those units could be spoiled just by their geography. There are five “incomplete” districts – Basarabeasca, Ocnita, Leova, Rezina, Soldanesti.

Rest 23 districts will be divided onto 3 categories.

First – districts with the full power control from DPM. It was no any electoral rivals for DPM in those districts. Thanks to some “preliminary efforts” from the DPM side. Therefore, it was no any limits to the fraud at this area. There are seven districts in this category, placed in the different geographical zone of RM.

Second. Those are districts with clear signs of gerrymandering and fraud just in the day of elections. Those led to the very tight wins for DPM with the minimal advantage. There are eight districts in this category.

Third. The districts when DPM lost for Acum or PSRM in SMC. There are also eight districts in category. PSRM won 5, Acum – 3.

We’ll note once more that there are districts and constituencies in the districts in Table 11. All three categories have no any geographical bias, each has districts from North, Center and South of RM.


Districts Results in districts by

party lists (mean), 2019

Swing for DPM
Acum PSRM DPM 2019 to 2014 single to proportional, 2019
Full power control for DPM 23% 26% 35% +18% +6%
Gerrymandering and irregularities for DPM 20% 31% 30% +12% +4%
DPM lost 27% 27% 28% +11% +3%

Table 11. Swings for DPM in districts from 2014 to 2019 and within 2019.


  1. Table 11 confirms the well-known formula. 30% is not enough to win in SMC with three real competitors, 35% is almost enough. This very thin margin gives just absolute result in our case. DPM won eight of eight in the category 2 and lost the same in the category 3.
  2. Wherein there is no great difference between swing in categories 2 and 3. If, it would be some different circumstances, DPM could capture some more “wins” in some other districts and constituencies from category 3. We can be sure, that if it would be the anticipate elections in the September 2019, DPM would proceed just in this direction.
  3. DPM supplied extra swing at 6% in category 1 from proportional system to SMC. We just noted that it was no real control in those districts, but combine of falsifications was launched. DPM gained +24% of votes in this category compared with 2014 which is 2.5 more than five years earlier. Therefore, DPM remained only third in the election rating. Such results totally contradict any statistical patterns and could be only fraud consequences.


Now we’ll turn to the “inverse” problem. We’ll study all those swings, but only in the boundaries of SMC. We turn back to the Table 10. DPM swings are more than 10% only in two categories from seven. There are less than 5% in three categories. Average from those five categories (less than 10%) – 4.2%. Further, we’ll consider only 4-5% growth for DPM as regular one. All totals, overpassing 5%, we will consider as irregularities and falsifications.

From the beginning, we’ll verify normality of SMC and exclude some SMC that are not normal in our study case. C17 (Nisporeni), c18 (Orhei), c19 (Orhei, Ivancea) would be immediately excluded for previous reasons. Additionally we exclude c4 (Riscani) – candidate from Acum at the last moment has withdrawn from elections and all votes of Acum (by proportional) came for DPM in SMC. C43 (Cahul) – candidate of DPM (only case) was elected as independent, and there is any extra swing between votes for DPM by list and in SMC. We excluded only 5 SMC from 30 and that is the minimal range. Rest 25 SMC we will split into three categories as it was in the previous paragraph.

There are nine SMC in the categories 1 and 3,  seven in the category 2. Totals look like as it was in the Table 11.

Constituencies Results in constituencies by

party lists (mean), 2019

Swing for DPM
Acum PSRM DPM 2019 to 2014 single to proportional, 2019
Full power control for DPM 26% 21% 37% +16% +5%
Gerrymandering and irregularities for DPM 19% 32% 31% +13% +4%
DPM lost 26% 30% 29% +11% +4%

Table 12. Swings for DPM in constituencies from 2014 to 2019 and within 2019.


  1. We just confirm the pattern. 35% is almost enough for win, 32-33% – not obviously in the SMC with three main competitors.
  2. Category 2 presents the main interest as the target one. PSRM leads in this category by total of votes and simultaneously leads in 6 of 7 SMC by votes on proportional system. Nevertheless, PSRM lost to DPM in all 7 SMC.
  3. Now we can calculate and define the “target swings” which were “designed” by DPM as the “planned falsifications”. There are 16-17% at the SMC compare with 2014 elections. DPM won all SMS, where it was fulfilled, and lost where not. Sometimes difference was the tightest one, but in most cases, it was under the “power control”.
  4. DPM won just “before” elections in the 8 districts and 10 SMC. Fraud at those districts were directed to the proportional. Falsifications in SMC were large one, but made “in addition” to proportional system voting.
  5. Therefore, we proved that 15% at areas were falsified by DPM and this swing was the part of the fraud. It took place in most districts and SMC. Sometimes DPM couldn’t provide the “target swing” (may be power control wasn’t so absolute), sometimes even 15% wasn’t enough (suddenly, especially in the towns, gaps were larger than anticipated). This 15% of “fraud swings” in areas were “seen” immediately after elections. Predominantly those preliminary calculations were fair. 16-17% means the 105-110 thousand of votes.




  1. We have to add 7-8% – 25 thousands in Chisinau with suburbs. Totally, there are 140 thousands of irregular votes in Moldova. As the result– 23.5% instead of 15% and five extra MP places for DPM by proportional.
  2. DPM won 7 from 18 SMC due to falsifications on the election day (c43 formally won the independent). There are c3, c5, c6, c16, c40, c41, c42.
  3. Most three SMC DPM won only by power control and pressure on the counter candidates.
  4. Gerrymandering (which was comprehensively studied at the previous chapter) and Election Day fraud were applied simultaneously to achieve the common effect. It is difficult to “split” the share of “each kind of fraud”.
  5. 1-2 SMC DPM won purely by tricks with geography and by tricks with the different number of voters in SMC.
  6. All kinds of fraud were prepared carefully, in detail and in depth. Some statistical irregularities were “hidden” under the specially corrected indexes. It is impossible to hide everything and mathematical and statistical analysis help us to discern most cases of irregularities and falsifications. Sometimes DPM couldn’t provide as it “was planned”. Especially it took place at cities and in the centers of districts. DPM had power control at some of those places, but social activists prevented some of this fraud.
  7. The final calculation demonstrates that DPM had captured extra 15-18 MP places from 31fornally obtained. Therefore, from the DPM part elections were falsified more than twice.
Share This