Possible reasons for inconsistent query results

I have a bunch of queries to be executed against an Access database as part of my Python script. Unfortunately, queries that are used directly in MS Access give some output records, in a Python script it returns nothing (no errors). Database connection and general syntax should be fine as simple queries (like select one column from a table somewhere) work fine. Here is the code for one of these tasks:

import pyodbc

baza = r"C:\base.mdb"
driver = "{Microsoft Access Driver (*.mdb, *.accdb)}"

access_con_string = r"Driver={};Dbq={};".format(driver, baza)
cnn = pyodbc.connect(access_con_string)
db_cursor = cnn.cursor()

expression = """SELECT F_PARCEL.PARCEL_NR, F_PARCEL_LAND_USE.AREA_USE_CD, F_PARCEL_LAND_USE.SOIL_QUALITY_CD, F_ARODES.TEMP_ADRESS_FOREST, F_SUBAREA.AREA_TYPE_CD, F_AROD_LAND_USE.AROD_LAND_USE_AREA, F_PARCEL.COUNTY_CD, F_PARCEL.DISTRICT_CD, F_PARCEL.MUNICIPALITY_CD, F_PARCEL.COMMUNITY_CD, F_SUBAREA.SUB_AREA
FROM F_PARCEL INNER JOIN (F_PARCEL_LAND_USE INNER JOIN ((F_ARODES INNER JOIN F_AROD_LAND_USE ON F_ARODES.ARODES_INT_NUM = F_AROD_LAND_USE.ARODES_INT_NUM) INNER JOIN F_SUBAREA ON F_ARODES.ARODES_INT_NUM = F_SUBAREA.ARODES_INT_NUM) ON (F_PARCEL_LAND_USE.SHAPE_NR = F_AROD_LAND_USE.SHAPE_NR) AND (F_PARCEL_LAND_USE.PARCEL_INT_NUM = F_AROD_LAND_USE.PARCEL_INT_NUM)) ON F_PARCEL.PARCEL_INT_NUM = F_PARCEL_LAND_USE.PARCEL_INT_NUM
WHERE (((F_ARODES.TEMP_ADRESS_FOREST) Like ?) AND ((F_AROD_LAND_USE.AROD_LAND_USE_AREA)<?) AND ((F_ARODES.TEMP_ACT_ADRESS)= ?))
ORDER BY F_PARCEL.PARCEL_NR, F_PARCEL_LAND_USE.SHAPE_NR;"""

rows = db_cursor.execute(expression, ("14-17-2-03*", 0.0049, True)).fetchall()

for row in rows:
    print row

cnn.close()

      

        
        
        
      

    

I know that these queries were generated from the query builder in MS Access, so I was wondering if this might be causing the differences, but on the other hand, it is still an access database.

Anyway, it seems that the problem is in SQL, so I would like to know what elements can lead to different conclusions between queries executed directly in MS Access and the pyopbc connection?